I'm looking for help with determining the cause of vibration in a generator. It is a 112mw air cooled unit attached to a gas turbine. This has been an ongoing problem for over a year. Initially we had high shaft vibration so balance moves were attempted. The shaft vibration was decreased after a number of moves but then we had casing vibration problems and axial vibration. Alignment was check and it was fine so they stiffened the generator by adding support plates from generator bearing to casing and added internal plates. This help some but over time the vibration has got worse. Recently if you change lube oil temperature or generator air temperature by a few degrees, the vibration would change. Now if you put any vars load on the generator vibration increases also. Exciter current has also increased by 50-75 amps higher than we normally see. Any ideas?
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Generator vibration cause 4
- Thread starter Rich246
- Start date
-
2
- #2
112 milliwatts - that is a small generator!!
More information is needed to figure out what is really happening.
Changing lube oil temperature could indicate an oil whirl type problem, you would need to look for the appearance of vibration frequencies between 40% and 49% of running speed.
As the problem seems to be getting worse with time, you could be seeing some sort of deterioration in the rotor construction (loose wedges, short ciruit etc).
More likely though is a thermally sensitive rotor: you can balance it to a very fine tolerance either in a workshop or in-site but when you apply load, the rotor heats up and 'bends' causing sypmtoms of unbalance - run the machine again cold or at no load and you find a well balanced rotor.
What you really need to do is some basic research such as:
a) Observe vibration during a cold startup and a hot shutdown and compare using a Bode plot format
b) run the machine up from cold and monitor 1X amplitude and phase for a sufficient period of time to establish that these values reach a steady value
c) add load and repeat the monitoring of 1X amplitude and phase angles until the reach a steady value - the vector difference between c) and b) represents the thermal contribution to the overall vibration.
d) repeat b) and c) to establish repeatability (of phase angle and amplitude)
Using the data from b) and c) you could attempt a compromise balance where, instead of using vectors from cold or no load conditios, you attempt to balance out the hot or loaded 1X vectors (or go for a compromise, where you aim for acceptable vibration at both no load and on-load conditions.
It can take several hours for vibration on a generator to stabilise follwing a load change.
More information is needed to figure out what is really happening.
Changing lube oil temperature could indicate an oil whirl type problem, you would need to look for the appearance of vibration frequencies between 40% and 49% of running speed.
As the problem seems to be getting worse with time, you could be seeing some sort of deterioration in the rotor construction (loose wedges, short ciruit etc).
More likely though is a thermally sensitive rotor: you can balance it to a very fine tolerance either in a workshop or in-site but when you apply load, the rotor heats up and 'bends' causing sypmtoms of unbalance - run the machine again cold or at no load and you find a well balanced rotor.
What you really need to do is some basic research such as:
a) Observe vibration during a cold startup and a hot shutdown and compare using a Bode plot format
b) run the machine up from cold and monitor 1X amplitude and phase for a sufficient period of time to establish that these values reach a steady value
c) add load and repeat the monitoring of 1X amplitude and phase angles until the reach a steady value - the vector difference between c) and b) represents the thermal contribution to the overall vibration.
d) repeat b) and c) to establish repeatability (of phase angle and amplitude)
Using the data from b) and c) you could attempt a compromise balance where, instead of using vectors from cold or no load conditios, you attempt to balance out the hot or loaded 1X vectors (or go for a compromise, where you aim for acceptable vibration at both no load and on-load conditions.
It can take several hours for vibration on a generator to stabilise follwing a load change.
Sounds like a couple problems, probably a load vector as mentioned by TPL, and it may be that the support stiffness in this generator is such that you really need to balance using absolute data, (casing vector added to shaft vector).
In any even, sounds like you need an in depth analysis of the problem as TPL described. Someone needs to take high density multichannel vibration data over both speed and load.
Where are you? Depending on location I may be able to recommend someone close by.
-The future's so bright I gotta wear shades!
In any even, sounds like you need an in depth analysis of the problem as TPL described. Someone needs to take high density multichannel vibration data over both speed and load.
Where are you? Depending on location I may be able to recommend someone close by.
-The future's so bright I gotta wear shades!
-
1
- Thread starter
- #4
Thanks for the help. I should add we have a Bentley system that reports to a computer and someone from our performance group is going over the data. The generator 13.8kv and just recently the vibration has got much worse. Most of the time load run steady for entire 24hr period and is changed only a few Mw from day to day. Load has much less effect on vibration than vars. Vibration before last week were around .34 in/sec and .3in/sec on casing vibration and highest shaft vibration was under 4 mils. Last week with vars on machine after a fuel swap the casing vibration was .491 in/sec and .41 in/sec with over 5 mils on highest shaft. Before last week vibration wasn't great but bareable usually below alarm levels of .37 in/sec. Now with any var load the vibration is in .44 in/sec range with highest shaft over 5 mils. I believe the change has do do with rotor shorting or stator end turn vibration since the vars seem to really effect it now when in the past it really had no effect. Does it sound like my beliefs are correct? I'm located on Long Island NY.
Yeah, if swinging vars has that much effect it is a hot spot. The Bently data should tell the tale, but again, it may be that you need casing data at the same time so you can look at absolute vectors. If the vector is repeatable, it is possible to balance out this effect for now, at least until it gets worse.
Contact the Bently office (now GE) in Philly, they have one or two real good vibration analyst left there...
-The future's so bright I gotta wear shades!
Contact the Bently office (now GE) in Philly, they have one or two real good vibration analyst left there...
-The future's so bright I gotta wear shades!
-
1
- #6
- Thread starter
- #7
- Thread starter
- #9
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question
