We need urgent help in form of vibration analysis and diagnostics of a very production critical multistage centrifugal compressor, driven by a steam turbine with speed increasing gearbox in between. The gas handled by the compressor is carbon dioxide which is first compressed in low pressure casing and then in this high pressure casing. The layout of the complete compressor train is attached as Annexure-I.
The overhauling of this compressor casing was performed to address the issue of high acid number of lubrication oil due to expected ingress of gas as result of high end seal losses. In this overhauling, rotor was reused along with existing stage separation seal (Material; Stainless Steel). Inter-stage (Material; Aluminum), end seal (Material; Aluminum) and bearings were replaced with new ones. Bearing clearance was set 0.11 mm (90 x 31 mm: LOP). The operating speed of the compressor is 13200-13800 with 1st critical speed as 7700-9800 rpm. Rotor total leghth:1401 mm and bearing span is 1027 mm The cross-sectional drawing of the compressor is attached as Annexure-I.
On 18th Jan, 2022 during start-up, the compressor tripped twice on high vibration of NDE radial bearing VT-205B and it touched 108 micron. Vibration on all other probes also increased but remained less than alarm (75 micron) and danger limit (100 micron). After internal discussion, it was decided to go a 2nd attempt by lowering the trip limit to 80 micron and reducing time delay from 01 second to 03 second. The compressor tripped in the same fashion and at this point, it was decided to perform detailed internal inspection Vibration signatures of both these trippings are available in form of M/s. Baker Hughes System-1 are attached as Annexure-II.
On dismantling the compressor bundle, the following two salient observations were noted;
- Rub between rotor (nearly 30% circumference) and stator (Inter-Stage and stage separation seal). All inter-stage seals are of aluminum and stage separation seal is of stainless steel with integral teeth and inter-locking configuration. The same is shown in pictures.
- Liquid accumulation in sealing gas piping, vent piping as well as in compressor internals.
- The observations of both overhauling’s are attached as Annexure-III.
Our findings for System-1 vibration signatures is;
- Referring Bode plot of 1st tripping, there was significance difference between ramp-up and coast down in form of direct vibration, 1X and phase angle. At the same speed of the shaft, this difference is more than 80 micron in VT-205B and it kept decreasing during the coast down.
- Referring the same bode plot of 1st tripping, there was a speed variation of more than 1000 rpm in less than 10 seconds and it repeated in nearly 2.5 cycles before tripping.
- During 2nd tripping, the ramp-up vibration again matched the ramp-up of 1st tripping by showing that rotor was exposed to temporary unbalance during the 1st The rotor responded with significant phase change (more than 90°) at VT-205B during the critical speed range of 7700 - 9800 rpm whereas no such behaviors exist in 1st trippings. After a few minutes, the vibration of compressor again increased with phase change but reverted back and finally it tripped on high vibration of VT-205B of 80 micron.
- In both these trippings, we were in loading phase of the compressor where recycle valve was being closed and compressor was being pressurized to meet the system requirement.
Your expert opinion is requested for the following queries;
- What are possible cause of both these trippings events?
- Is the liquid accumulation in a centrifugal compressor can cause temporary unbalance and sudden speed variation?
- If thermal bow generated by rub is the cause of high vibration, what is the plausible explanation for source of rub?
- Is any symptoms of mechanical fit-up issue are present in the vibration signatures?
- Machine Layout and Drawing as Annexure-I
- Sysytem-1 Vibration signatures as Annexure-II
- Observations of 1st and 02nd Overhauling’s as Annexure-III.