We are facing a persistent and unresolved issue with an 1850 kW ID Fan motor (VFD-driven), where the Non-Drive End (NDE) bearing repeatedly fails within ~3 months of operation under load.

Problem Description

• The failure is always preceded by progressively increasing vibration levels.
• This issue occurs irrespective of the motor used (i.e., even after replacing the motor with another healthy one).
• When a new or overhauled motor is installed, vibration levels are initially normal.
• After ~3 months of operation, vibration starts increasing gradually and ultimately results in NDE bearing failure.
• Once the motor is replaced, the same pattern repeats.

Bearing Details

• DE side: NU240 C3
• NDE side: NU236 C3 + 6236 C3
• Bearings are from reputed manufacturers, and no inherent bearing defects are observed.

Vibration & Diagnostic Observations

• FFT analysis indicates housing looseness.
• However, physical inspection does not reveal any looseness in the bearing housing or foundation.

Actions Already Taken (Over the Years)

1. Motor foundation bolt holes enlarged to eliminate potential soft foot issues.
2. Laser alignment carried out between motor and ID fan.
3. Dynamic balancing of ID fan recently performed to address mass eccentricity.

Despite the above corrective actions, the failure trend remains unchanged.

Key Concern
Why does the motor consistently run vibration-free initially and then develop increasing vibration leading to NDE bearing damage within ~3 months, regardless of the motor installed?

Request for Discussion
We are seeking expert opinions on:

• Possible root causes (mechanical / electrical / system-related)
• Role of VFD-related effects (bearing currents, resonance, thermal growth, axial loading, etc.)
• Any missed checks or advanced diagnostics recommended
• Long-term corrective measures to permanently eliminate this issue

Any insights, similar experiences, or suggested investigation approaches would be highly appreciated.

 

Have the bearings checked for pitting

Read through this

https://www.helwigcarbon.com/understanding-circulating-currents/

 

Thank you for the response and for sharing the article — it was very informative and helped clarify the circulating current mechanisms.

To add some details on our existing setup: the motor does not have an insulated bearing on the ODE side, however the ODE bearing housing itself is insulated by design, and the measured IR value is ~17 MΩ. In addition, a shaft grounding ring is installed on the ODE side.

As mentioned in the article, shaft grounding rings are often recommended on the DE side, whereas in our case the ring is on the ODE side. Interestingly, we have not observed any DE bearing failures so far; the repeated failures are still confined only to the NDE bearing.

I have attached a photo of the ODE side NU236 C3 bearing inner race for reference:

• Right side: inner race after removal from site
• Left side: new bearing inner race

One observation is a slight circumferential groove on the used inner race, indicating that the rollers may not be in full contact across the entire raceway width, as the wear pattern is confined to only a portion of the inner race width. I would appreciate views on what could cause this kind of partial-width contact pattern — whether it points more towards load distribution, axial effects, or electrical phenomena.

As a next step, we are planning to measure shaft-induced voltage during motor running to assess the actual effectiveness of the insulated bearing housing and the shaft grounding ring in service conditions. I will share the findings once measurements are taken.

Looking forward to further insights and suggestions.

 

For VFD:
- what input voltage is and what kind of connection to grid (direct, 1 or 2 transformers)?
- what switching frequency is set?
- how long VFD_motor cable is? Any output filter?