Hi

We are running two 500kva gensets with an paralleling panel which is only used to transfer the load between the generators when one of them is due for maintenance or a repair (They run 24/7). The power is supplied to a 500kva rated VFD which drives an submersible oil well pump. The system has been running with no problems since it was installed 2 years ago, two months ago the pump had to be pulled out from the well because of an mechanical fault, a new one was lowered in as a replacement, now that a new pump is installed the well needs to go back into production so... we start up one of the generators, closes the breaker and the VFD operator starts up the pump, as soon as the pump starts to pull power the generator and the control panel goes crazy, synch controllers starts puling down the RPM on the generator tills the generators trips on "Under Frequency", using my multimeter to measure the frequency, instead of getting 50HZ i get 150HZ so think to my self this most be harmonics, so an harmonic test was done and the 5th harmonic has gone up to 64% which i can't explain why, we are still using the same Generators, Control Panel, VFD, Step up Transformer, with the same power consumption as before, the only difference is a new exact copy of the old pump installed replacing the old broken one.

So what do you guys think could cause this? to me this is strange, even the guy in charge of the VFD say's that he haven't changed any setting on the VFD.

 

Sorry... I forgot to mention that the load always runs on 113KW 162KVA. the load is stable, but it does some times go up a few kilowatts for a few seconds and then fall back down to 113KW, but that is normal and rarely happens.

 

SWAG:

When the pump motor failed, it pulled a lot of current from the VFD and caused internal damage. something like one of transistors cracking and partially leaking to ground or maybe one of the DC bus caps burning through. The damage isn't enough (yet) to trigger a protection circuit in the VFD, but is increasing the DC bus ripple so much that it affects the rectification, which in turn results in higher than normal harmonics.

I'd get a service tech from the VFD mfr out ASAP to do a thorough PM call on it while it can still be saved, because if the failure cascades, you will lose the whole thing, maybe the motor again with it...

 

Some things to think about:

*Any chance any instrumentation got damaged during the pump changeout - speed or phase probes, current voltage insturments.

* Any chance wiring got reterminated incorrectly.

* Any fuses or circuit breakers blown?

* Did you try starting the motor using the other genset? Can the gensets be run in parallel prior to a pump start?

* Was the damage to the pump verified to be the cause. What damaged the pump? Are you sure you didn't have 2 problems, and one still remains?

* Motor rotation is correct?

* Any capacitor cans bulged or leaking? Capacitors can usually be checked without much disassembly.

* Any signs of overheating damage anywhere?

* Does everything work properly with the genset running prior to starting the pump? Are there a set of baseline readings in this condition that you can compare with?

* Does the pump rotate freely?

Sort of sounds like the motor might be slipping poles during the start process. What did the actual speed on the motor do during the start? Did it almost get to full speed before it died - did it not accelerate at all and just draw excessive current until it tripped - or something in between?

If the simple checks don't produce anything and you don't have access to high level troubleshooting equipment it is probably best to call the OEM and have a tech check things out.

 

Starting one or both generators and paralleling them together can be done with no problems and they operate fully normal "while the VFD is OFF". I've done a resistive load bank test on the system with no problems so the suspect is for sure the VFD.

The pump was replaced due to a broken shaft.

I do have a range of test instruments available, i'll try to test the DC bus voltage for AC ripples and the caps and see what results i get. Speaking in general considering VFD's... what is the limit for AC ripples on the DC bus ?.

 

I don't have a number for you on the normal or allowable AC ripple on the DC bus. Perhaps there are other VFD experts out there that can provide some typical numbers. I assume you have thoroughly reviewed the O&M manual for all information that might help you diagnose the problem. It would be best to look at the waveform. The ripples should all be uniform. if one out of every 3 or every 6 ripples is substantially different, that might point you to a failed capacitor, fuse, or connection.

It is always good in a complex system such as a VFD to take a set of baseline voltage readings and if possible waveforms for as many circuits as possible - when things are working as they should before something goes wrong. The readings should be at several different load configurations. These will come in handy later on when things aren't working.

Most likely the pump shaft broke because of long term fatigue, hopefully after many years of operation. The break should show evidence of beach marks indicating propagation of a small crack over a long period of time until the cross section became too small and then broke the rest of the way in one shot. Any chance the shaft broke because of something drastically wrong.

A VFD should produce equal torque on the motor from each of the 3 phases, so that the sum of the 3 phase torques is constant. If one phase produces considerable more or less torque than the others, there will be a pulsating torque on the pump shaft - much like a combustion engine. This could be detrimental to the pump shaft life. Checking the voltage and current on each phase to the motor should be done as a minimum. Having the voltages and currents from all 3 phases plus the DC bus hooked up to a high speed recorder (> 2K samples/sec) would be ideal for troubleshooting. Then you could see what is going on during a startup. Test equipment can be rented for this.

Some other ideas are to check for unwanted grounds in the system by performing megger checks. Thermal imaging cameras can also be used to look for hot spots and high resistance connections.