Case Studies: Practical Findings of Measuring Loop Inspections
The importance and informative value of turbomachinery vibration measurements is beyond dispute. The following list of deficiencies demonstrates that a regular examination of any vibration measurement system is virtually indispensable:
- Faulty transmitters on two new systems; only approx. 20% of the actual values were indicated. The settings of the GAP signals were absolutely correct. Everyone was pleased about the low vibration values, but the compressor was not monitored! All of the transmitters were exchanged at the cost of the manufacturer.
- The damage to the shielding on an extension cable reduced the loop sensitivity to < 5 mV/µ; therefore, the indicated and monitored values were too low by approx. 40%.
- On an approx. 18 year old monitor, the loop sensitivity of all sensors was below 6 mV/µ; all indicated and monitored values were too low by at least 25%.
- One monitor was configured to Smax evaluation; the limit values, however, were parameterized in accordance with a peak-to-peak evaluation and were, therefore, too high by a factor of 1.5.
- There is a growing trend toward transmitter solutions with limit value monitoring in the control system. The zero point of some vibration transmitters proved instable and occasionally fell short of 4 mA, which regularly led to false wire breakage messages. In order to stop the constant fault messages, the wire break monitoring feature was deactivated. Values < 4 mA were set at Zero. When a wire break then actually occurred and the transmitter emitted 3.5 mA, this was consequently not recognized as a wire break. The Zero indication was interpreted as good running smoothness in conjunction with an inaccurate transmitter. The compressor was not monitored!
- In the case of vibration monitors, a defect probe can be recognized on the OK message; on a majority of the examined systems, the OK message was not processed by the control system or only processed as a fault message. Therefore, the machine was not monitored during wire break!
- On a gearbox-driven compressor, a 9m system was built onto the gearbox and a 5m system was built onto the compressor. When, during a bearing test, two extension cables were destroyed and needed to be replaced, they were replaced by cables of the wrong length, which subsequently resulted in wrong values.
- If differing vibration systems are installed at a plant, there is a high risk that non-harmonized or unsuitable components are combined with one another. Such combinations mostly issue indications that can be significantly distorted.
- In a case in which a changeover to a replacement transmitter with a deviating measuring range was carried out, this factor was not taken into account in the control system. This has, in fact, led to expensive and dangerous wrong decisions in cases involving 3 different systems:
- An insufficient measuring range results in a higher value and consequently leads to a machine shutdown. Due to the fact that the measurement was trusted, first the bearings were checked and then the rotor was dismounted and balanced at high speed. It was only during the course of re-commissioning, after several unnecessary weeks of downtime, that the wrong shaft vibration transmitter was discovered.
- An insufficient measuring range results in a higher value and, with a higher load, consequently leads to a machine shutdown. Because of this load-dependency, the machine was only operated with a reduced load for several months, until the next planned standstill, which resulted in a significant loss of production.
- During the exchange of all transmitters, the old, considerably lower, measuring range setting was kept in the control system. After the revision, the operators were delighted about the tremendously improved vibration behavior, but the machine was insufficiently protected.
- Interference of poorly laid measuring cables that are parallel to current-carrying lines.
- In the case of ground loops, a 50 Hz vibration is typically superimposed over the actual vibration signal. This can distort the vibration level and lead to a completely wrong interpretation of the cause of vibration, especially with operating speeds of 25 or 50 Hz.
- After 5 mm probes got replaced with 8 mm probes from another manufacturer, the measurement failed. It turned out that the lateral gap of the probe was too small due to the modified probe diameters.
- High vibration with pronounced portion at 2 * n have been temporarily explained with bearing play. After checking the bearing though, it was obvious that the tag number did not fit to the measuring point and the wrong bearing had been considered.