APFC Malfunctioning On DG Set

So the APFC unit is sensing voltage and no current, and stays out of the picture? That sounds like correct behaviour.

No, APFC is receiving voltage from the BUS. Sensing current is only from the Mains load.

Infact, by not sensing the current from common bus eliminates the problem of high voltage due to higher capacitive reactive load, when DG is running.

But, in our case, due to all the capacitors being switched ON, when DG is running, voltage rises to 480V !

But, after the inductive load is started again, voltage drops, but capacitors do not switched off !

You have a classic "race" condition where two controllers are trying to control a single point.

Here's what's happening, when you run your DG in isolated mode the AVR attempts to maintain the voltage as the generator takes on load, at the same time your APFC sees the increasing load and "thinks" to itself "...Ah-ha, the load is increasing therefore I have to add some capacitance to improve the power factor..". But what is the result of adding capacitance?...the voltage goes up! And what does the AVR do when it sees the voltage go up above its set-point? It tries to lower the voltage even further, which causes the APFC to see a worsening power factor so it switches in more capacitance. This will continue until all 14 stages have been switched in, or the generator trips on under-excitation.

Consider yourself lucky that the generator didn't become so severely under-excited that its air-gap magnetic field became so weak that it could no longer effectively transmit the mechanical power from the prime mover to the load thereby causing governor hunting, instability, and ultimately slipping poles and tripping. Worst still would be if the added capacitance just happened to reach a point where it was equal to the inductance of the generator, resulting in a resonant condition. Things would then accelerate at electronic instead of mechanical speeds. These are a few of the reasons why most DG set manufacturers advise against running their generators below 0.95 pf under-excited.

Fortunately the fix is easy, simply operate the APFC in manual whenever you are not connected to the grid.

I have a confusion !

I have stated that the sensing CT is kept in the Mains out going Bus, not in the common bus. So, when mains is off, CT sensing will be zero irrespective of the load on common bus, from DG set. In this case, all the capacitors in the bank should always be off when DG is running.

The same is working alright in all the previous installations commissioned so far, ranging from 140 kva to 720 kva installations. But, this problem has arised in the latest installation.

What is the reason ? can any body explain?

I have a confusion also. How does an APFC (or any controller for that matter) operate properly when it is missing one of its input variables? The instructions for this device:

http://www.lntebg.com/esp/Service/Downloads/Manuals/ServiceManual/Installation_Troubleshooting_Capacitors_APFCSystem.pdf

make the assumption that the CT is measuring something otherwise the control panel will indicate "1- -0" but does not tell us what action the controller takes under these circumstances..

Stating that the controller operates properly in other different installations does not tell us that it will work properly in this one because as you have stated they are not identical.

Have you performed the basic troubleshooting steps mentioned in the manual? Have you ensured that the controller is even operating properly? Doesn't it seem odd to you that the controller automatically switches in all the capacitors when it has no way to calculate the power factor since it does not know the current.

The first thing I would do is get everything running and loaded with the controller in manual, verify that the pf indicated on the panel is correct, and then switch to automatic. If that fails then I would swap out the controller board for one known to operate properly. If that fails then go to one of your other installations and short out the CT connection and see how the controller behaves.

Please do let us know the results of your tests instead of simply repeating your question.