3 pole transfer switch

[don_resqcapt19]

If I have an industrial installation with no line to neutral loads and a 3 pole transfer switch can I install a system bonding jumper in the transformer and a second one in the generator and run supply side bonding jumpers from both the transformer and the generator to the transfer switch? The feeder conduits between the transfer switch and both sources of supply will have three 500 kcmil and a 1/0 system bonding jumper in each conduit.

 

[Smart $]

You can, but if you're going by 2011/2014 NEC, be aware your GEC connections have to be in the same enclosure as the system bonding jumper for separately derived systems [250.30(C)]. Also, that 1/0 conductor is called the supply-side bonding jumper (SSBJ).

 

[Smart $]

Just want to make sure here...

Are these systems wye configured and you are bonding the neutral terminal but not running a neutral conductor to/through TS? ...or are they delta configured and corner grounded?

 

[don_resqcapt19]

Just want to make sure here...

Are these systems wye configured and you are bonding the neutral terminal but not running a neutral conductor to/through TS? ...or are they delta configured and corner grounded?

Yes, a grounded wye system but no line to neutral loads. The grounding electrode conductor connections and the system bonding jumpers will be installed in the transformer and at the generator. The supply side bonding jumpers will be run with the phase conductors from each source to the transfer switch.

 

[GoldDigger]

Since the neutral is not being switched, regardless of the position of the transfer switch you have two ground to neutral bonds in widely separated locations, both after the service disconnect. That is not allowed.
The two options under the NEC are to configure the generator as a NON separately derived system and omit the ground to neutral bond at the generator. You run both a neutral and an EGC to the generator, and you can connect a local ground electrode to the EGC, but you cannot bond neutral to EGC.
The alternative is to bond in both places but switch the neutral so only one bond is in play at any time.

 

[Smart $]

Since the neutral is not being switched, regardless of the position of the transfer switch you have two ground to neutral bonds in widely separated locations, both after the service disconnect. That is not allowed.
The two options under the NEC are to configure the generator as a NON separately derived system and omit the ground to neutral bond at the generator. You run both a neutral and an EGC to the generator, and you can connect a local ground electrode to the EGC, but you cannot bond neutral to EGC.
The alternative is to bond in both places but switch the neutral so only one bond is in play at any time.

Hmmm... I assumed the transformer was a separately derived system, not a service transformer. With only one SDS connected at any one instant, there is technically only one system bonding jumper in that instant.

 

[GoldDigger]

Hmmm... I assumed the transformer was a separately derived system, not a service transformer. With only one SDS connected at any one instant, there is technically only one system bonding jumper in that instant.

If all you are switching with the transfer switch, there are always two bonds, not counting the service bond, on the circuit which is the output side of the transfer switch.
The fact that one of the bonds is associated with an offline system does not make it any less a bond.
If you put a GF device on the generator output, it will trip as soon as you connect a load, because of the bond at the transformer.

This issue comes up most often with portable generators, whose manufacturers feel that UL requires them to include the bond as part of the generator wiring. Not always easy to disable.

 

[Smart $]

If all you are switching with the transfer switch, there are always two bonds, not counting the service bond, on the circuit which is the output side of the transfer switch.
The fact that one of the bonds is associated with an offline system does not make it any less a bond.
If you put a GF device on the generator output, it will trip as soon as you connect a load, because of the bond at the transformer.

This issue comes up most often with portable generators, whose manufacturers feel that UL requires them to include the bond as part of the generator wiring. Not always easy to disable.

Butttttt... with no neutral conductor and no line-to-neutral loads, there will be no line-to-neutral faults (at least none outsides the source enclosures), and you have to provide a line-to-ground fault current path. You can't do that with two grounded and switched separately derived systems and only one SBJ.

The other option is as you stated, make one system non-separately derived... but then you have to run a [grounded] neutral conductor to/from both sources while still having to run SSBJ's and a GEC to at least the separately derived source.

Would you choose to do the latter when you can compliantly do the former?

 

[ActionDave]

I hate the alphabet soup the NEC has come up with, SBJ, SSBJ, MBJ.....etc., when all we need to do is make sure there is a way to clear a fault and there is only one bond at the service or at a separately derived system.

When I read the OP I thought the same as Golddigger, bond at the transformer and make the generator a non-sds.

 

[GoldDigger]

, ...and you have to provide a line-to-ground fault current path. You can't do that with two grounded and switched separately derived systems and only one SBJ.

If you do not run a neutral as far as the transfer switch from both sources, you do not have a double bond. If you do run a neutral from both sources to the transfer switch, then either bond by itself will provide the needed fault current path for both.
I am not convinced that the NEC allows you to omit the neutral connection between either source and the disconnect, whether there are line to neutral loads or not, as long as the xfer switch is the first disconnect, but you may be able to persuade an inspector of that.

 

[iceworm]

Don -
Is the system 480V or 208V? If 480, both the gen and the xfm are 750KVA? The transfer switch is 1200A? Doesn't matter much, mostly curiousity - I want to get a feel for the system.

If so:

That's bigger that most any transfer switch I've ever seen. I'd guess it is made up of two motor operated CBs. And the cbs, of course are 3pole. Most I have seen are two power CBs in a switchboard - electrically interlocked, maybe sync equipment for a seamless transfer. Still the physics should be the same.

My inclination is to run both feeds to the transfer switch as 5 wire. Since there are no single phase line-to-neutral loads, the neutral can be pretty small. Collect al the neutrals together. Collect all the EGCs together. Make one N-G bond in the transfer switch.

Edit to add: you may also want to put in a neutral disconnect on both the xfm and the gen. Good idea to have this for maintnenace.

If the xfm is utility owned and the utility insists on an NG bond in the xfm - I'll have to rethink this.

ice

 

[don_resqcapt19]

This is all on the plant side of the service. The system is 480y/277 solidly grounded, but with no line to neutral loads. The transformer is 2000 kVA and the gen set 2000kW. Eight sets of three 500's with 1/0 supply side bonding jumpers between the two sources and the ATS.

The transfer switch is rated at 3,000 amps. It is a GE/Zenith switch, but not sure of its construction.

I think that the installation is a technical violation of the NEC with grounding electrode connections and system bonding jumpers at both the transformer and the generator, but since there are no line to neutral loads, I don't see a real world issue.

 

[don_resqcapt19]

If you do not run a neutral as far as the transfer switch from both sources, you do not have a double bond. If you do run a neutral from both sources to the transfer switch, then either bond by itself will provide the needed fault current path for both.
I am not convinced that the NEC allows you to omit the neutral connection between either source and the disconnect, whether there are line to neutral loads or not, as long as the xfer switch is the first disconnect, but you may be able to persuade an inspector of that.

I am not aware of a requirement to run the grounded conductor to the first disconnect for an SDS. It is there for a service, but not there for an SDS.

 

[GoldDigger]

I am not aware of a requirement to run the grounded conductor to the first disconnect for an SDS. It is there for a service, but not there for an SDS.

Code details aside for the moment, I think I can come to terms with a situation in which neither SDS brings a neutral to the transfer switch (and no neutral leaves the transfer switch, of course), but I have a problem with an asymmetric situation where one SDS supplies a neutral and the other does not.
If there is no neutral at the transfer switch, you really don't have to switch it!

 

[Smart $]

...

I think that the installation is a technical violation of the NEC with grounding electrode connections and system bonding jumpers at both the transformer and the generator, but since there are no line to neutral loads, I don't see a real world issue.

Compare to the same system where you run neutral conductors and use a 4-pole TS. Each system would be still be SDS. Under 2008 and earlier codes, you'd make the GEC and SBJ connection at the TS one the load side. Under 2011/2014 code, the SBJ and GEC connection(s) is(are) required to be at the SDS source. So you'd have an SBJ and GEC connect at both sources.

With no LN loads, no neutral conductor, and a 3-pole TS, how does this make it a technical violation under 2011/2014 codes?

 

[don_resqcapt19]

Compare to the same system where you run neutral conductors and use a 4-pole TS. Each system would be still be SDS. Under 2008 and earlier codes, you'd make the GEC and SBJ connection at the TS one the load side. Under 2011/2014 code, the SBJ and GEC connection(s) is(are) required to be at the SDS source. So you'd have an SBJ and GEC connect at both sources.

With no LN loads, no neutral conductor, and a 3-pole TS, how does this make it a technical violation under 2011/2014 codes?

I still see it as the neutrals bonded at two locations via the supply side bonding jumpers.

 

[Smart $]

I still see it as the neutrals bonded at two locations via the supply side bonding jumpers.

True... but there are no LN loads and no neutral conductors, so there'll be no concern for neutral current parallel paths on grounding conductors.

 

[GoldDigger]

I think that the original statement of the problem caused some of the confusion, in that it left open the possibility of an unswitched neutral at the TS rather than that there simply were no neutral conductors present and no direct connection at all between the solidly grounded neutrals at the two sources.

 

[Smart $]

I think that the original statement of the problem caused some of the confusion, in that it left open the possibility of an unswitched neutral at the TS rather than that there simply were no neutral conductors present and no direct connection at all between the solidly grounded neutrals at the two sources.

I agree, it could have easily been read that way. I made the assumption there were no neutral conductors being an industrial installation which are typically 3?, and only three circuit conductors in each conduit. Could have also been 1? with switched neutrals. Your consideration for possibilities was essentially the same reason I followed up with configuration questions in Post #3.

One thing I'm wondering about is whether there are disconnecting means on the supply side of the TS. Generators typically have an integral breaker, but transformers do not.

 

[GoldDigger]

:thumbsup:
No line to neutral loads is not quite the same as no neutrals, although one might tend to imply the other. But the sources as described were not actually deltas.