1MW VSD failure
1MW VSD failure
(OP)
We experienced a 1MW variable speed drive failure recently.
The drive is connected to 6.6kV swithboard and 24-pulse type. The input power supply transformer is epoxy cast type and is the one failed. It resulted in 6.6kV Y&B phase fuse failure (6.6kV fuse backed latched Vacuum contactor) and happened when switched-on. The make is Ansaldo, France. Investigation is on-going.
Any thoughts welcome!!
The drive is connected to 6.6kV swithboard and 24-pulse type. The input power supply transformer is epoxy cast type and is the one failed. It resulted in 6.6kV Y&B phase fuse failure (6.6kV fuse backed latched Vacuum contactor) and happened when switched-on. The make is Ansaldo, France. Investigation is on-going.
Any thoughts welcome!!





RE: 1MW VSD failure
There are several reasons for the elevated kV/mm level. One is if there are higher frequency harmonics. The high frequency increases capacitive leakage currents and that, in its turn, increases ozone production. The other reeason is the overvoltage that occurs when you switch open the vacuum contactor. If you do not have a voltage limiting device (arrester or RC combination) the overvoltage will almost allways be destructive over time.
I have had the latter phenomenon in quite a few places. The first one was around 1974 with the first vacuum contactors used for MV motors and the latest was a couple of years ago. Surge arresters helped in all cases.
The former phenomenon was not so easy to find. Normal harmonics analysis disn't show anything abnormal. But using an oscilloscope with FFT revealed high levels of non-harmonic resonance voltage on the 6.6 kV side. That HF voltage had the same effect and produced ozone in the voids. Three transformers exploded from that reason before we could find the reason. The Italian manufacturer is now defunct and we changed to Geafoil insulation. Those transformmers are still working very satisfatorily.
In both cases, it took 9 - 15 months before the damage occured.
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
The subject drive is a new one and is still under commissioning. It is true the 6.6kV feeder to the converter trafo is switched by Vacuum contactor.
But, the dry type converter trafo is provided with Surge arresters at its 6.6kV terminals (and the surge arresters do not seem to have suffered any damage, testing of SAs is yet to be done).
RE: 1MW VSD failure
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
Then it was switched-off for making cable connection from the inverter output to the motor.
After the connection, when switched-on, the failure occurred (thyrister firing is off).
The failure was in the middle (vertically speaking) of the winding and shows rupture of winding insulation.
RE: 1MW VSD failure
It is quite common that the transformer is damaged at switch off because of the fast action of the vacuum bottles. But nobody notices until the next switch on when the damage causes a flash over or other breakdown.
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
I will let you know once the surge arrester is tested.
RE: 1MW VSD failure
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
Further, if the failure is related to voltage surge, you would expect the damage at the line end and not in the middle of the winsdings as is the case here.
Could it be winding failure due to quality issues (the transformer is still new and under commissioning as stated earlier)!!
The 24-pulse VSD package is from Ansaldo, France and the transformer is from Trafomec, Perugia, Italy.
Has anyone got experience with the above makes!
RE: 1MW VSD failure
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
RE: 1MW VSD failure
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
One was put down to a certain lot of insulating material employed in between the turns. The transformer was repaired and mica based insulation material used instead in the new coils and no further problems. The transformer manufacturer should be able to trace lots of materials and advise where they have been used in yr equipment and cross-check with other customers.
The second was with a 12 pulse transformer that was in parallel with another for a 24 pulse drive.
One unofficial comment to me (not from the transformer manufacturer though) was that as the transformers had extended delta and/or zig-zag windings to get the appropriate phase displacements, (I can't remember which now) the likely cause of the failure was a manufacturing problem within the windings connections.
Damage is definitive in a cast resin transformer so the actual coils will not be repairable and it might not be possible to really determine what happened. How long the fault lasted will determine collateral damage perhaps to the core?
RE: 1MW VSD failure
I guess it could be design / manufacturing competency as the transformer for a 24-pulse VSD is much complicated.
It could also be material / manufacturing / quality problem if moisture ingress in to the cast resin insulation had really happened.
Only a detailed investigation could bring out the fact. A 3rd party specialist is expected for the investigation.
Skogsgurra,
You mentioned Gonella make transformers. I could not find their website. However, it seemed that both Trafomec and Gonella are Italian companies.
RE: 1MW VSD failure
Gunnar Englund
www.gke.org
--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.
RE: 1MW VSD failure
RE: 1MW VSD failure
On high fault rated contactors , high speed fault clearance is sometimes achieved by the use of very powerful springs.
Bleed off Air piston type damping cylinders are used to dissipate excess energy at the end of the stroke .
I've experienced that sometimes the size of the piston bleed off orifices are incorrectly sized for the altitude (denser air)& contact bounce occurs.
This causes high frequency high voltage transients & consequent weakening & damage to transformer windings, insulation,capacitors & motors.
You need to perform oscillographs of the action (movement) of the contact mechanisms when opening, to establish if this is your problem.