Separately derived systems

[petersonra]

So the other day I was thinking about a portable system I was working on. It has 480 volt 3-phase coming in that is transformed down to 208-120 for the actual equipment. If this was an SDS, some kind of grounding electrode connection would be required. This is moderately inconvenient for a portable system.

So it occurred to me that if I were to connect the neutral of the secondary side of the transformer to the neutral of the 480 coming in, it would no longer be a separately derived system and thus would not require a grounding electrode.

This seems almost too simple of a solution to me so I'm wondering if I have missed something here.

Being as the neutrals will be connected together eventually via the EGC, I don't see a downside electrically or safety wise, and I'm not convinced that having an extra grounding electrode serves any real purpose.

I haven't had time to look real close at the requirements for such things so maybe I am missing something obvious.

 

[synchro]

So it occurred to me that if I were to connect the neutral of the secondary side of the transformer to the neutral of the 480 coming in, it would no longer be a separately derived system and thus would not require a grounding electrode.

One thing to note is that the neutral conductor on the 480 side would have to handle the L-N short circuit current coming from the secondary of the transformer during a ground fault.
I assume this is a delta-wye transformer?

 

[electrofelon]

One thing to note is that the neutral conductor on the 480 side would have to handle the L-N short circuit current coming from the secondary of the transformer during a ground fault.

Please elaborate. I'm not seeing it.

 

[synchro]

One thing to note is that the neutral conductor on the 480 side would have to handle the L-N short circuit current coming from the secondary of the transformer during a ground fault.

Please elaborate. I'm not seeing it.

Because the neutral to EGC bond is in the 480 system and not in the 208 system, a ground fault current on the 208 side would flow from the neutral point of the 208Y secondary, through the connection to the neutral conductor of the 480 feeder, and over to the N to EGC bond of the 480 system. The amount of ground fault current available would be the L-N short circuit current at the secondary of the transformer. The impedance of the 480 neutral conductor and EGCs would reduce this current somewhat.

 

[LarryFine]

Draw it out and follow the fault-current path. Remember, fault current must reach the secondary.

How would a line-to-ground fault on the secondary find its way back to the secondary's neutral?

Also, the primary's neutral must be large enough to serve as every conductor you're asking it to.

 

[electrofelon]

Because the neutral to EGC bond is in the 480 system and not in the 208 system, a ground fault current on the 208 side would flow from the neutral point of the 208Y secondary, through the connection to the neutral conductor of the 480 feeder, and over to the N to EGC bond of the 480 system.

I was assuming there is a N-G bond at the secondary as per normal. The problem with that is you are connected neutral to ground past the first disconnect in violation of 250.24(A)(5). If there is not a N-G bond at the secondary, then......I'd have to think on it some more

 

[petersonra]

I was assuming there is a N-G bond at the secondary as per normal. The problem with that is you are connected neutral to ground past the first disconnect in violation of 250.24(A)(5). If there is not a N-G bond at the secondary, then......I'd have to think on it some more

There would be no neutral to ground connection on the secondary of the 480 to 28 0 transformer. That was one of the things I was thinking of as being a requirement for an SDS which might kill this idea altogether. How do people deal with this problem? It can't be something that no one has ever run into before?

 

[electrofelon]

There would be no neutral to ground connection on the secondary of the 480 to 28 0 transformer. That was one of the things I was thinking of as being a requirement for an SDS which might kill this idea altogether. How do people deal with this problem? It can't be something that no one has ever run into before?

Seems to me without a N-G bond and both supply and load neuters connected to XO you have a setup that essentially looks like an autotransformer. Of course this has two windings unlike an auto, but I don't see any requirement for one winding to make it a non SDS. I cannot come up with a code violation.

 

[synchro]

Seems to me without a N-G bond and both supply and load neuters connected to XO you have a setup that essentially looks like an autotransformer. Of course this has two windings unlike an auto, but I don't see any requirement for one winding to make it a non SDS. I cannot come up with a code violation.

I agree that at least the neutral connection is similar to using an autotransformer.
Speaking of autotransformers: petersonra could you use a 3-phase autotransformer such as the one at the link below instead of a delta-wye for your application? This one is a 15 KVA but they make a range of sizes.

http://equotes.hammondpowersolutions.com/cgi-bin/equote_edb.pl?pn=Y015QTCF3L0U

 

[don_resqcapt19]

I would install the system bonding jumper on the 208 side and connect it to the primary EGC in leiu of supplying a grounding electrode and GEC. Yes, I am aware that this would not be in compliance with the actual code language, but don't see a lot of options for a portable system and don't see safety issues.

 

[electrofelon]

I would install the system bonding jumper on the 208 side and connect it to the primary EGC in leiu of supplying a grounding electrode and GEC. Yes, I am aware that this would not be in compliance with the actual code language, but don't see a lot of options for a portable system and don't see safety issues.

There is the option of using the EGC as the GEC and having it be compliant if the conductor meets the requirements of both. Not sure if that is possible with what Mr Peterson is doing.

 

[don_resqcapt19]

There is the option of using the EGC as the GEC and having it be compliant if the conductor meets the requirements of both. Not sure if that is possible with what Mr Peterson is doing.

I would not even worry about it being suitable for use as both for this application. There is nothing magic about the connection to earth, and I have no issue using the primary EGC as the GEC for this type of application, even though the code does not permit it. If I were the AHJ, I would give written permission using the provisions of 90.4. I just don't see a safety issue.

 

[winnie]

The legalistic/code understanding of an 'SDS' is one thing, the physics is another.

As far as the physics is concerned, a 480 delta to 208/120 wye transformer is a separately derived system. You have a source of current flow that has no electrical connection to another source of power. This is important because any fault current must return to its source, in this case the transformer secondary. It doesn't matter if the source of 208/120 is a diesel generator, a motor/generator set or a magnetically coupled transformer secondary. Current will try to flow back to the 208/120 source.

I tend to agree with Don about the lack of magic in a connection to the soil. The only potential issue that I see is that any EGC type connection to the metallic surroundings of this piece of equipment should be sized for the possible fault current of the 208/120 secondary. IMHO the EGC in the going to this portable piece of equipment should be sized based on the OCPD on the 208/120 side of things.

-Jon