Test Motors, through its online predictive maintenance tool, TestIM, allows detecting the faults of electric motors or generators and their components, in this case the bearing of an alternator. It is the generator of a cogeneration of a power of 3.7 MW
During the month of January 2018, the TestIM system automatically detects an increase in vibration on the drive side of the generator, in the vertical direction.
The RMS vibration values are deviated with respect to the values that the system estimates should be, for the operating conditions that the machine is having. This estimation is a consequence of the "Machine Learning" learning process, which the TestIM_SmartSoft system develops. We can see it in the first graph, the blue curve against the green curve. The vibration oscillated between 80-20m/s^2, when it was appropriate to keep it at a level of 30m/s^2.
According to the algorithms developed by Test Motors, this increase in vibration is due to a lubrication deficit.
The graphs show the increase in RMS vibration values during the month of January. We can also see how these values greatly exceed the values established by "Machine Learning" system, between the end of January and the beginning of February.
According to the recommendation of Test Motors, the customer lubricated the bearings for several days. After greasing, the vibration behavior in the generator has improved considerably, reaching the vibration values established by the "Machine Learning" system (30m / s ^ 2) as detailed in the following graph.
In this way, the incident is given by finished.