The CO2 centrifugal compressor (HP) occasionally has spikes in the trends of shaft vibrations on both bearings. There is a 0.83X subsynchronous component in the spectrum, the magnitude of which changes over time. Additional subsynchronous vibration components also appear at irregular time intervals. Their frequencies are 1/3 and 2/3 of the 0.83X component. The problem of rotating stall in compressor impellers is suspected. Attached are some plots (trends, spectrums, orbits).
Dear Riro ,
The occurrence of this transient phenomenon depends on certain operating conditions and loop configurations of centrifugal compressors. The temporary rotor bow may cause uneven internal clearances ,which can be produce sub harmonic peaks below 1X Vibration .
I am not an expert in such huge Turbo machinery Analysis ,but i am posting my views .
Your graphic does not have the data that you relate. Do you know if the 1/3 and 2/3 are exact fractions or are approximate?
New machine, old machine, after repair, any changes to process - pressures or temperatures.
Yes, the 1/3 and 2/3 are exact fractions of 0,83X component. The machine is old one, almost 40 years in operation. There were no changes in process parameters. For your information the 0,83X component has always been present in the spectrum. After the last repair of the compressor (replacement of the thrust bearing) in May 2020, these spikes appear in trends and 1/3 and 2/3 of the 0,83X fraction are visible in the spectrum.
The plots look like DM2000 (superseded by System1). Very good to see the system working but better have a plan for supporting it if you don't want to upgrade.
On the analysis, I would not say anything before knowing the speeds of the high speed and low speed shafts of the gearboxe. Is it a single stage?
Can you confirm if the high vibration is only observed on this point 22-XE-108-1?
Regards- Ali M. Al-Shurafa
The DM2000 will be upgraded in the near future.
As you guessed, gearbox speed increaser is single-stage. Both speeds are visible in the first attached plot: 6924 rpm and 13454 rpm. The vibrations are high on the remaining 3 proximity probes as well (22-XE-108-2, 22-XE-109-1 and 22-XE-109-2).
I did quick calculations but couldn't relate these frequencies to something clear. I suggest to repost the pdf after you add the frequencies to the peaks to save the time for the others.
I would trend the data with the process parameters such as suction flow rate, discharge temp, recycle valve opening etc.
DM2000 has a feature that you might want to use: high resolution data. The spectrum will be processed using 3200 line, if I remember correctly.
Regards- Ali M. Al-Shurafa
The speeds, 6924 rpm/13454 rpm yield a potential tooth combination of 35 x 68. Nothing in that combination that would generate the frequencies in question.
Can you confirm the 35 x 68 tooth combination?
Bull gear: 101 teeth; Pinion: 52 teeth