800 ASq D VFD Cabinet

[ivsenroute]

Has anyone worked on this particular setup?
It appears as though the B phase on this 277/480 system let loose between the breaker and fuses. It tripped the breaker above and the feeder to this cabinet from the electrical room. I am just curious as I never worked on this particular unit before and was curious about the setup. I believe it is about 15 years old.

 

[Cow]

Had any thermal imaging been done at regular intervals before this happened?

It appears that B phase may have been running hot for a while as evidenced by the B phase fuse possibly being replaced once already.

 

[Besoeker3]

Maybe a bad/loose connection at the breaker?

 

[kwired]

Have to agree with some of others, likely connection issue on B terminal of the breaker, and the fact the corresponding fuse appears is different brand, maybe has blown before and was a sign something isn't right.

Best advice is change the breaker as well as the fuse holder (and the wire between) at very least, or it may do it again, they may both be compromised now.

 

[chris kennedy]

What are the devices below the fuses?

 

[Besoeker3]

What are the devices below the fuses?

Look like chokes to me.

 

[Jraef]

The first thing after the fuses is a line reactor, aka "AC choke", then below that is an 18 pulse transformer, used to help try to mitigate the harmonics. It's going to have 3 phases in, 9 phases out in three sets of three phases, each one phase shifted by (usually) 20 degrees. That then feeds 18 diodes all configured as 3 separate bridge rectifiers, then those all feed a common DC bus for the VFD. The 18 pulse diode bridge is over to the right side under the two fuses. Here's a general schematic, although it doesn't have the line reactor ahead of it (or the DC link fuses).

One thing I would suspect in your panel is that those cables from the breaker to the fuse block appear to be extra flexible cables, meaning a high strand count, probably Class K stranding or higher. That kind of cable requires special compression lugs and dies for the tools. The bottom lugs (attaching to the fuse block) and those on the reactor and transformer appear to be factory machine made, but are different from those attaching to the breaker. So I would suspect that at the breaker end the compression, lug, die or all three were incorrect or incorrectly done, leading to an overheated connection and eventual melting of the conductor. I've seen that too many times; people don't understand that the high flex cable needs special connections.

I don't know if that was a Sq. D factory built 18 pulse drive or not, even though all the parts are Sq. D. Part of my reason for doubting it is that someone put fuses on the DC link, which is not something a VFD mfr. typically does, it's something that is done by someone who is being overly cautious.

 

[Besoeker3]

The first thing after the fuses is a line reactor, aka "AC choke", then below that is an 18 pulse transformer, used to help try to mitigate the harmonics. It's going to have 3 phases in, 9 phases out in three sets of three phases, each one phase shifted by (usually) 20 degrees. That then feeds 18 diodes all configured as 3 separate bridge rectifiers, then those all feed a common DC bus for the VFD. The 18 pulse diode bridge is over to the right side under the two fuses. Here's a general schematic, although it doesn't have the line reactor ahead of it (or the DC link fuses).

One thing I would suspect in your panel is that those cables from the breaker to the fuse block appear to be extra flexible cables, meaning a high strand count, probably Class K stranding or higher. That kind of cable requires special compression lugs and dies for the tools. The bottom lugs (attaching to the fuse block) and those on the reactor and transformer appear to be factory machine made, but are different from those attaching to the breaker. So I would suspect that at the breaker end the compression, lug, die or all three were incorrect or incorrectly done, leading to an overheated connection and eventual melting of the conductor. I've seen that too many times; people don't understand that the high flex cable needs special connections.

I don't know if that was a Sq. D factory built 18 pulse drive or not, even though all the parts are Sq. D. Part of my reason for doubting it is that someone put fuses on the DC link, which is not something a VFD mfr. typically does, it's something that is done by someone who is being overly cautious.

Yes. We have done a few multipulse systems but we didn't put the bridges in series. It doesn't provide the most efficient system and optimum efficiency can win you the contract. Or lose it if you don't.

 

[RumRunner]

. . . I don't know if that was a Sq. D factory built 18 pulse drive or not, even though all the parts are Sq. D. Part of my reason for doubting it is that someone put fuses on the DC link, which is not something a VFD mfr. typically does, it's something that is done by someone who is being overly cautious.

_____________________________________________________________________________

But, could that redundant protective feature a contributing factor to the failure?
From what I can tell there is no evidence of runaway heat buildup in the DC link.

I would think a sustained (slow development ) of loose connection on the middle phase.

 

[kwired]

The first thing after the fuses is a line reactor, aka "AC choke", then below that is an 18 pulse transformer, used to help try to mitigate the harmonics. It's going to have 3 phases in, 9 phases out in three sets of three phases, each one phase shifted by (usually) 20 degrees. That then feeds 18 diodes all configured as 3 separate bridge rectifiers, then those all feed a common DC bus for the VFD. The 18 pulse diode bridge is over to the right side under the two fuses. Here's a general schematic, although it doesn't have the line reactor ahead of it (or the DC link fuses).

One thing I would suspect in your panel is that those cables from the breaker to the fuse block appear to be extra flexible cables, meaning a high strand count, probably Class K stranding or higher. That kind of cable requires special compression lugs and dies for the tools. The bottom lugs (attaching to the fuse block) and those on the reactor and transformer appear to be factory machine made, but are different from those attaching to the breaker. So I would suspect that at the breaker end the compression, lug, die or all three were incorrect or incorrectly done, leading to an overheated connection and eventual melting of the conductor. I've seen that too many times; people don't understand that the high flex cable needs special connections.

I don't know if that was a Sq. D factory built 18 pulse drive or not, even though all the parts are Sq. D. Part of my reason for doubting it is that someone put fuses on the DC link, which is not something a VFD mfr. typically does, it's something that is done by someone who is being overly cautious.

Only thing I have to say is the burned off insulation on the B phase, seems to be revealing typical 19 strand conductor from what I can see...

Still looks like the crimp connection is likely what failed for some reason though.

 

[Jraef]

But, could that redundant protective feature a contributing factor to the failure?
From what I can tell there is no evidence of runaway heat buildup in the DC link.

I would think a sustained (slow development ) of loose connection on the middle phase.

I didn't mean to imply that the DC link fuses were a contributing factor, only that it is evidence of something more likely done by a panel builder, not the Sq. D factory.

Only thing I have to say is the burned off insulation on the B phase, seems to be revealing typical 19 strand conductor from what I can see...

Still looks like the crimp connection is likely what failed for some reason though.

DLO cable in that size would have multiple bundles of fine strands that often look like typical cable at first glance and it's hard to tell the difference via photos. But having seen multiple failure of this sort, that's what it looks like to me.

 

[chris kennedy]

There was a resolution to this thread. The B phase conductors were replaced and the VFD came back to life, wasn't there, don't know about finely stranded or not.

Here's the rub.

This is a water treatment facility for the city of Highland Beach. If you work in the Palm Beach County area and are interested in 24/7 365 emergency service for this city (mostly lift station stuff) PM me.

 

[ivsenroute]

Thanks for all of the replies. I appreciate the input. Great information.