Negative VARS

Juan, a few questions about your situation:

What is the generator size in relation to the others?

Is there a grid connection or is the generator being synchronized with other similar sized units?

What type of loads are there, resistive, motors, and/or power factor correcting capacitors?

When you say that the "...generator seems to slowly walk away from the bus voltage...", at what rate is this happening?

What is the control strategy for the other generators, are they also on PF control?

Without knowledge about how your system is configured it is hard to pinpoint the nature of your problem, but a few things come to mind:

If there are other similarly sized units and they are operating properly then I would verify that the polarities of all the CTs and PTs providing inputs to this unit are verified to be correct.

I would also check the voltage and PF of the load before this unit is synchronized, if the voltage starts rising after synchronization then it could be an interaction between the other units and this one, and when this unit sees the bus voltage rising it will naturally try to lower it by reducing the field current, the cycle continues until the underexcitation or undervoltage protection disconnects the generator.

Also check that the generator is set for parallel operation and that all the other generators are set and wired properly as well.

If there are PF correction capacitors make sure that any automatic power factor controllers (APFC) are not sensing the same point as this generator. If they are electrically close to each other then it is possible that you will have two controllers trying to control the same variable and you will have a condition where one "wins" at the expense of the other; i.e., as the APFC adds capacitors (and raises the voltage) the generator APFC sees the rising voltage and lowers the excitation in an attempt to keep the terminal voltage from rising.