delta to delta transformer grounding

[pgordon]

I have a client who is requesting a transformer: 500kVA, 480V delta primary, 240V delta secondary. The transformer will be supplied from a 480/277V, 3-phase, 4 wire switchboard (neutral to ground bond in the switchboard) and will exclusively supply a DC plant with non-isolated rectifiers, 240V, 3-phase input, no 120V loads, no neutral required for the loads. The output of the DC plant is going to be 380VDC, with a high resistance midpoint ground (HRG) which will limit the line to ground voltage to 190VDC and the HRG will limit the ground fault current. The client has requested a delta to delta transformer because they do not want a neutral to ground bond and ground electrode installed on the AC side of the DC plant. Their concern is that the solidly grounded wye output would defeat the current limiting characteristics of the HRG on the DC side of the plant, causing a worker to be exposed to higher ground fault current.

My question is, whether or not we are required by the NEC to ground the delta to delta transformer as a separately derived system, or can we just rely on the equipment grounding conductors as a continuous path back to the main 480/277V, 3-phase, 4 wire switchboard?

 

[winnie]

You cannot depend upon the EGCs back to the original 480/277 supply to carry fault current/trip breakers/etc.

When you have a transformer (delta-wye or delta-delta) you introduce electrical isolation that prevents current flow. If you don't ground the secondary somehow, then you have an ungrounded system; with all of the issues that go along with such.

Grounding the neutral of the DC bank does accomplish the necessary grounding to permit fault current to flow.

Resistance grounding the neutral of the DC bank provides the benefits of resistance grounding (controlling system voltages like a grounded system, reducing fault current like an ungrounded system.)

But I don't think this approach would fit the letter of the NEC. No-where in article 250 can you resistance ground DC systems, nor AC systems <480V. IMHO this approach is a good idea, and it would make sense to get engineered plans drawn up and permitted by the AHJ rather than trying to follow the letter of the NEC.

-Jon

 

[Smart $]

I recall reading somewhere for grounded DC plant that has no isolation from its AC source, the AC supply is considered grounded... or maybe it was the other way around... but I can't it now for either way.

That said, I have no expertise in this area and have never heard of an HRG DC system. Perhaps someone more versed will chime in...:happyyes:

E2+: I see Jon posted while I was researching and typing, so your in good hands...

 

[winnie]

you might consider a resistance grounded wye secondary instead of a resistance grounded DC string. The DC string would have a bit of common mode AC ripple, but faults from the AC side of things would not have to go through the rectifiers.

Additionally, if the sole reason for resistance grounding is to reduce fault current (rather than for continuity of service) you could consider using ground fault relays to trip breakers in the event of faults (not GFCI for personal protection, but GFPE to trip breakers when these resistance limited ground faults occur.)

-Jon

 

[Smart $]

you might consider a resistance grounded wye secondary instead of a resistance grounded DC string. The DC string would have a bit of common mode AC ripple, but faults from the AC side of things would not have to go through the rectifiers.

Additionally, if the sole reason for resistance grounding is to reduce fault current (rather than for continuity of service) you could consider using ground fault relays to trip breakers in the event of faults (not GFCI for personal protection, but GFPE to trip breakers when these resistance limited ground faults occur.)

-Jon

Even with no expertise in this area, I agree from a perspective of overall understanding of electrical systems.

 

[Bugman1400]

I don't quite understand the reasoning behind the customers of the OP. On the DC side, the midpoint is grounded like any DC battery system would be in order to detect ground faults on a DC system. An HRG on the DC will not limit the DC current flow when the other leg also comes into contact with ground. What can limit the DC current flow is the rectifier unit itself. Since the DC is not a DC battery and is a rectifier with electronics perhaps there is circuitry that can detect a DC ground fault and limit the DC current before the rectifiers are overloaded. This is also typically done in the reverse case with DC inverters where the inverters will limit the DC current flow to twice the rated output. There are no AC protection devices (GFCI, AF, etc.) that can detect ground faults on the DC system from the AC side.

On the AC side, as Don, alluded to, there are a whole set of other issues to contend with ungrounded systems like ensuring your equipment is L-L rated instead of L-G rated.