TVSS exploded

[niceyuan]

Dear All:

Another problem:

As a result of the damaged s/s, we did rewired our system to accomodate injection of 4 rented 480V gensets to our existing switchgears. Our existing system has 4-4160v feeders from s/s supplying 4 low voltage switchgears through 2mva xformers (1 for each feeder. Each switchgear with installed TVSS 250 kA, 480V; Each switchgears also with capacitor bank to correct our pf. When we removed the cables from the main of LVSG and installed the supply cables from rented gensets our TVSS exploded,,,all three of them...What seems to be the cause? our suspect might be harmonics? resonance? need your technical suggestions.

Thanks

 

[waross]

If you connected the capacitors before connecting the load, the leading power factor may have over excited the generators, leading to overvoltages.
respectfully

 

[resqcapt19]

Was the grounding set up on the output of the generators the same as what was on the secondary of the transformers?
Don

 

[jraef]

If you connected the capacitors before connecting the load, the leading power factor may have over excited the generators, leading to overvoltages.
respectfully

Ahhh, one of the joys of bulk correction...

Just as a side note, if you have and VFDs that were connected to the line at that time you might want to check them before re-energizing again. Also if you had any soft starters you may want to check them as well in case they had MOVs on the SCR circuits.

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[alehman]

When you disconnected the 480V source, did you by chance disconnect the neutral-ground bond? Or was it at the transformer? Some TVSS's can't tolerate N-G voltage.

 

[JBinCA]

I would expect that resqcapt19 and alehman are on the right track and that your grounding scheme changed.

If your system is 480 / 277, the TVSS is comprised of MOVs that were sized for the 277 voltage (plus some design margin of 15% or so). When appling MOVs to any other grounding scheme than solidly grounded wye, the proper voltage to size them by is the line to line voltage of 480 (plus the design margin).

The MOVs conduct over their strike voltage and are probably rated to conduct the 250 ka (which seems high to me) for something on the order of 10 miliseconds.

Jraef and waross, what would the typical magnitude of the over-excitation be, and for what duration?

Regards, JB

 

[waross]

As the power factor goes leading the system (capacitors) supplies magnetizing vars to the generator. The more the power factor goes leading the less excitation is provided by the Automatic Voltage Regulator. A system with bulk power factor correction may have enough capacity to provide more than normal excitation current without any input from the AVR.
This is not a transient condition. The voltage may increase to the point that further increases are limited by saturation of the generator iron. It may persist until some load is added to absorb some of the leading VARs, some capacitors are disconnected, or something fails.
I would guess in the range of 20% to 50% over voltage. Possibly more. Comments invited, jraef.
The use of a standby generator is one of the site conditions to consider when evaluating the suitability of bulk power factor correction. It is well to consider automatically disconnecting part or all of a bulk PF correction scheme when going on generator power.
respectfully

 

[JBinCA]

Waross,

Great info, thanks! I did a little searching to find a reference that explains this phenomenon and came up empty handed. Do you know of an online refernce you could point me to?

Niceyuan,

The phenomenon waross describes would blow your TVSS to kingdom come.

Did anyone note what voltage was indicated between the time of closing the generator breaker(s) and the TVSS failure?

Regards,

JB

 

[niceyuan]

Thanks for the replies:

1. When we removed original feeders and inserted gensets supplies the same setup was utilized we have y secondary with ground and neutral but we have no neatral load.

2. WE have VFDs and other harmonics contributor loads but on the downstream distribution panels we have another set of TVSS installed.

3. NO abnormal voltages noticed (we have nexus revenue meters installed recording also harmonics. During the energization I noticed an unusual high Voltage total distortion up to 34%. But on one occasion i am monitoring on one LVSG a low reading of distortion but then the TVSS suddenly exploded. On all explosions all 4 wirings of TVSS burnt and exploded also.

4. on 1st explosion all our connections from gensets have no grounding, then we installed grounding on the remaining connections but still they did explode.

Thanks

 

[coconutalley]

I used to design surge protectors with MOVs for massive
computer market. MOVs are only good for spikes and cannot tolerate sustained overvoltage. They can be rated for fairly high peak current e.g.250kA but one should note that it is for a spike of say 8us x 20us or 10us x 1000us. Based on your symptoms [and experience on getting returns] tells me that the MOVs saw overvoltages that is longer than the above. MOVs will catch on fire or explode if its ratings are exceeded. They typically fail short and become a filement for the AC power. Check the TVSS breakover voltage and this should tell you what level it will trigger at. Also, calculate the energy. You'd want TVSS breakover voltage to be above the peak of the 480VAC + tolerance [highest normal swell expected] to be safe.