Jrva,

is the intermittent peak happen during normal operating condition of 24 MW? is there a previous maintenance happen on this machine? 

can you share the plots, when did your last startup happen? can you share also its plot.

to your post "Your expert opinion is required"... even an expert probably cannot solve your problem without any data to see and analyse.

regards,

RG

Dear Rotating Guy,

Thank you for your response.

Yes it happens during normal operating condition. Start-up of machine was carried out in March, 2018 and start-up data is not indicating any abnormality. These events started after June 20, 2018.

Last year machine was tripped multiple times due to high vibration on generator. Subsequently, generator was sent to OEM where it was inspected and it was revealed that retaining ring at drive end of generator was loose which resulted in short winding. It was repaired at OEM shop and design of retainer ring was improved. Afterwards, machine vibration data did not show any abnormality till June 20, 2018.

By looking at the polar plot and spectrum, one would think it may be a case of intermittent rub (reference attached). However, it is the periodic behavior of the event that bothers me i.e. event is repeating itself after ~ 7 hours and in my experience repetition of intermittent rubbing would not be that much precise.

Best Regards,

Jrva,

could share the trend plot of this event? Is the vibration increasing when your unit at steady state normal operating condion? Could you check the voltage gap of the sensor? (In my history if experience -16 gap for prox probe leads to chattering and make the trend shows some intermittent peak).  I wouldnt say that your problem is on mechanical, if i am at your position then i would clear first E&I problem that may occur in your trend as you observe. Also, how many seconds thus the intermittent spike happen? Thus stepchange happen during intermittent spike?

please share your data.

regards,

Dear RG,

Plots have been attached for your reference.

Maximum amplitude of vibration during these events is not increasing with time at steady state normal operation. However, it has been observed that frequency of occurrence of these events is increasing.

No abnormality found in gap voltages.

During these events vibration increases gradually.

Best Regards,

I'm sure you have checked but could you share if anything else has changed?  Bearing temperatures, oil pressure, ambient conditions, etc.  To discount the Morton effect completely, try changing the oil temperature and see if this has an effect.  If you have access to an ultrasonic analyser, try taking the output from the ultrasonic unit into your vibration analyser and use the demodulator (Peak-Vue) function - that will more readily identify if there is a mechanical rub.  

Can you share the full spectrum waterfall also to identify the rubbing or load shifting issues.