A motor-pump-turbine alignment

Background
The drive train consists of a motor, pump, and recovery turbine and has been experiencing problems over the past few months . . .

Finding
From the laser alignment readings taken it was confirmed that indeed there was a severe misalignment between the components. The first set of readings taken was between the motor and pump (see graphic below). The motor was designated as the movable machine for this set of readings. The angularity and offset in the horizontal direction (0.9 mils/inch & 0.9 mils respectively) were determined to be outside the acceptable tolerances for a machine operating at 3580 rpm. Similarly, the coupling values in the vertical direction (angularity of 1.1 mils/inch and offset of 13.9 mils) were also well beyond acceptable tolerances.

It was noted that there were no shims under the feet of the motor. It was explained to us that because the motor was so massive and heavy, the pump was used as the "movable" machine when the alignment was I initially done.

Next, a set of reading was taken between the pump and the recovery turbine (the turbine was designated as the movable machine). This set of readings (below) also revealed that the misalignment condition between these two elements were also out of tolerance.

Since there was no facility for moving the motor, a set of machine train readings were taken with the motor designated as the stationary (the machine to the far left with zero values). This plot gave an overall picture of where the pump and recovery turbine were sitting with respect to the motor.

Below is the final condition of the pump with respect to the motor. The coupling values are significantly improved over the "as found" conditions. Further vertical movement was impeded by the turbine side coupling binding slightly as a result of the severe misalignment of the recovery turbine.

Below are also the results of the alignment of the recovery turbine to the pump. Again, the coupling values are significantly improved over the "as found" condition.

Following the alignment a set of vibration measurements were taken. These new sets of data showed that the characteristics of misalignment ad all but disappeared from the vibration spectra. However, there were now high amplitudes at 1X, as well as harmonics that usually indicate imbalance and or looseness.

The following day we checked the alignments again to confirm for the plant personnel that the alignments were acceptable. All the values repeated, so the decision was made to begin dynamic balancing.

Prior to performing the alignments, soft foot was checked. It was found that the motor had a soft foot of 5 mils that was corrected. On the pump we found a soft foot of 14 mils that was corrected.

There was some suspicion that the piping might be causing the pump casing to move once on-line. We installed the laser heads to the side of the pump casing to measure axial, lateral, and vertical movement as the machine was turned on. In each case we saw no appreciable movement.

Identical turbines - grow differently

Situation
A power plant on the east coast asked Vibralign, Inc to provide absolute thermal growth measurements on #5 and #6 LM-6000 Turbines. These machines, 2 among 10, have a history of high vibration during operation and alarm level vibration following condensing operations.

What we found
Although the machines were identical in design and duty/performance, some of them were not capable of achieving design loads due to excessive vibrations. The machines had been installed and aligned per the OEM's recommendations.

What we did
In order to obtain the required thermal growth data, we installed 2 laser-based monitoring systems. This allowed the dynamic measurement system to be installed inside the turbine enclosure and monitor the changes in the vertical position of the Inboard and Outboard supports for the turbine.

Results
The OEM recommends leaving the outboard end of the turbine 78 mils low and leaving the inboard end 18 mils low. The difference between the OEM and measured turbine outboard end growth is 32 mils. The difference between the OEM and measured turbine inboard end growth is 15.5 mils. The growth results were graphed over the OEM recommended cold alignment position of the machine. This allowed us to accurately calculate the actual operating alignment condition of this machine. The following represent the calculated operational alignment condition of Turbine 5: VO: -70 mils VA: +0.5 mils/1".

Based on the above calculated operational alignment condition of the turbine and the previously collected data between the clutch and the generator, the estimated operational angle of the gears in the clutch are as follows: Input Shaft Angle: 3.5 mils/1" Output Shaft Angle: 0.7 mils/1" Estimated Clutch Gear Angle: 4.2 mils/1".

The clutch gears are rated to operate at no more than 0.5 mils/1" under normal operating conditions for extended periods of time and no more than 1.0 mil/in for brief periods.

During the previous visit we noted high vibration levels on the permanently installed vibration sensors on the clutch following condensing operations of the unit. For this visit we used a portable vibration data collector to collect and analyze the vibration data on this machine during start up and at normal operating conditions.

The installed accelerometers are mounted near the bottom of the clutch housing and are oriented in the vertical direction. The vibration levels indicated on the monitor in the control room indicated 0.06 in/sec (peak) and 0.11 in/sec (peak) on the turbine end and generator end of the clutch, respectively. Vibration data collected on the bearing caps shows vibration levels much higher than those indicated by the installed accelerometers.

If the relationship between the installed vibration sensors and the measured vibration on the bearing is linear, when the machine is alarming at 0.6 in/sec the actual vibration on the clutch generator end bearing could be as high as 1.85 in/sec. At shutdown levels of 1.0 in/sec the actual vibration could as high as 3 in/sec.

While the thermal growth data is very similar between Units 5 and 6, the vibration data is dramatically different. While the vibration on Unit 5 relatively high in amplitude, the vibration data on Unit 6 shows a smooth operating machine.

Given that the changes in the alignment between the turbine and the clutch are very similar for units 5 and 6, another explanation for the differences in the vibration data is the initial alignment condition of Unit 5. The spectral data collected on Clutch 5 shows the dominant vibration occurring at 1X shaft RPM. This typically indicates an unbalance condition exists on the rotor. This type of gear arrangement may also show high levels of 1X Shaft RPM vibration as a result of extreme misalignment. Phase data was not collected to determine if this vibration is due to unbalance or misalignment, however the thermal growth data and the OEM recommended cold alignment targets indicate that a severe misalignment does exist.

Unit 5 operation vibration levels abnormally high for this type of machine. The dynamic measurement data collected on Unit 5 indicates a severe misalignment condition exists between the turbine and clutch. Correcting this misalignment will improve the operating condition of this machine.

Alignment of a 2 Stage Compressor

Basic procedure
Atlas-Copco compressor at Akzo Nobel Eka Chemicals
The application is a two stage compressor which consists of an electrical motor with two output shafts (one in each end), the motor is then coupled with two turbines, one on each output shaft.

The two turbines and the electrical motor are all together working as a two stage compressor. The motor has a rotational speed of 2900 rpm (1800kW) and it's connected to the turbines via two gearboxes. The gearboxes have an output speed of 8580 and 12576 rpm respectively and the rotor (turbine) weight is 600kg/each. The couplings between the motor and the gearboxes are flexible couplings. The speed and power involved are really putting the alignment in focus. And it was great to use the machine train program in order to avoid unnecessary movements of the machines. To use any other alignment method than laser would have been extremely time consuming but with Fixturlaser Shaft 200 it took us 4 hours to check and align the machines.