Reverse fed 480/277-2300 delta secondary

[Jraef]

That's what I said! It really depends on available fault current. If arc flash is a huge concern, I'd go ungrounded.

Eh... not entirely (or did I miss another comment?).

You can run ungrounded, but you'll need ground fault monitoring equipment definitely

Not definitely... that was my point. It's a good idea, but not a definite requirement per the NEC.

 

[xptpcrewx]

there is no requirement for Medium Voltage (1kV and up) to have a Delta system grounded or to have GF monitoring.

Unless you’re dealing with mobile/portable equipment…

Most of the Delta MV systems I have worked on are not corner grounded but do have protective relaying that would detect a GF and therefore would meet the same intent as 250.21.B.

What type of protective relaying for an ungrounded system are you referring to?


Sent from my iPhone using Tapatalk

 

[DrSparks]

Eh... not entirely (or did I miss another comment?).



Not definitely... that was my point. It's a good idea, but not a definite requirement per the NEC.

I didn't say it was. Pretty important though IMO.

Sent from my SM-A326U using Tapatalk

 

[xptpcrewx]

You need the neutral on the 480 side or you wouldn't have a complete circuit.

No, the fault current on an ungrounded system would not return to the xfmr. That's the whole point of an ungrounded system--fault current mitigation.

Sent from my SM-A326U using Tapatalk

This isn't actually the point of an ungrounded system. It's mostly installation cost and service continuity. Note: An ungrounded system is not immune from phase-to-phase faults (which is basically what happens when you fail to remove the first line-to-ground fault), so fault current mitigation isn't really a benefit.

 

[DrSparks]

This isn't actually the point of an ungrounded system. It's mostly installation cost and service continuity. Note: An ungrounded system is not immune from phase-to-phase faults (which is basically what happens when you fail to remove the first line-to-ground fault), so fault current mitigation isn't really a benefit.

I agree with that. Especially when service interruption could cause a greater hazard than the arc flash.

Good point on phase to phase fault, although less likely than a phase to ground fault.

Sent from my SM-A326U using Tapatalk

 

[xptpcrewx]

That's what I said! It really depends on available fault current. If arc flash is a huge concern, I'd go ungrounded.

Sent from my SM-A326U using Tapatalk

Making the system ungrounded doesn't really get you anywhere for arc-flash hazard mitigation either because personnel are still exposed to the incident energy resulting from phase-to-phase faults. As mentioned, the main benefits for ungrounded systems are installation cost and service continuity, but the disadvantages associated with this type of system can outweigh the benefits and is one reason it is discouraged.

 

[jminer99er]

Back to my one of my brain hiccups; How is a medium voltage unintentional ground fault on the steel structure ok with this low voltage system. (Egc on the secondary, that's bonded to the case, that's bonded to the xo, that goes back to the gear?) Wouldn't a ground fault make everything in that system act as corner grounded?

 

[winnie]

Current always try's to get back to it's source. Not to ground, not to other systems nearby, but back to its source.

In the setup described here, the source of the 2300V system is the transformer. There is no 'galvanic connection' from the 480V side to the 2300V side, so current won't try to flow from one side to the other.

A grounded phase conductor on the 2300V side _will_ 'corner ground' that portion of the system, but this won't change much on either 480V stretch. (There always is a bit of capacitive and inductive coupling, so no system is truly 100% isolated by a transformer, but it is pretty close.)

So in the event of a fault on your 2300V stretch your 480V stretches remain center grounded. You get a bit of current injected at the fault point of the 2300V system, this returns capacitively to other parts of the 2300V system. This small current is not enough to noticeably energize anything bonded to ground.

Jon

Jon

 

[kwired]

I agree with that. Especially when service interruption could cause a greater hazard than the arc flash.

Good point on phase to phase fault, although less likely than a phase to ground fault.

Sent from my SM-A326U using Tapatalk

If you don't have ground fault indication and are operating with a ground fault, the next ground fault on another phase can be basically about same thing as phase to phase fault.

 

[DrSparks]

If you don't have ground fault indication and are operating with a ground fault, the next ground fault on another phase can be basically about same thing as phase to phase fault.

Yeah I never thought of that. Isn't that what GFCI protection for equipment is basically for?

Sent from my SM-A326U using Tapatalk

 

[xptpcrewx]

Yeah I never thought of that. Isn't that what GFCI protection for equipment is basically for?

Sent from my SM-A326U using Tapatalk

No. As far as the NEC is concerned, GFPE is applied to solidly grounded wye systems exceeding 150-V to ground but not exceeding 1000-V phase-to-phase.

 

[xptpcrewx]

How is a medium voltage unintentional ground fault on the steel structure ok with this low voltage system.

The same way it's ok to bond the primary and secondary grounded conductors of every other system on the planet via EGC. Whether its one-end grounded, center grounded, corner grounded, mid-point grounded, etc. doesn't make a difference to the other side of the transformer.

(Egc on the secondary, that's bonded to the case, that's bonded to the xo, that goes back to the gear?)

EGC on the secondary, that's bonded to the case, that's bonded to the X0, that goes back to the gear, that dead-ends there and has no galvanic connection back to the MV system....

Similarly, you may consider EGC on the secondary, that's bonded to the case, that's bonded to the earth, that's connected to all other grounded transformers on the planet... What difference does it make?

Wouldn't a ground fault make everything in that system act as corner grounded?

Define "everything" and clarify what you mean by how a corner grounded system "acts"...

 

[kwired]

Yeah I never thought of that. Isn't that what GFCI protection for equipment is basically for?

Sent from my SM-A326U using Tapatalk

Won't work for an ungrounded system. First fault doesn't leak any current outside the protected path. It takes a second fault before it would trip.

 

[winnie]

I had the same concern, so I added a $3k protection relay that nobody wanted to pay for and had the answer beaten into me by my boss after he had to eat the relay cost...

Dumb question, but why so expensive?

It would seem fairly straight forward to detect if a 2300V phase had grounded, and then to shunt trip the supply 480V breaker. Would doing so add thousands beyond the required costs?

Thanks
Jon

 

[xptpcrewx]

Dumb question, but why so expensive?

It would seem fairly straight forward to detect if a 2300V phase had grounded, and then to shunt trip the supply 480V breaker. Would doing so add thousands beyond the required costs?

Thanks
Jon

You would need a PT and the gear to house it (aside from whatever protective relaying and integration costs necessary to implement it).

 

[jminer99er]

The same way it's ok to bond the primary and secondary grounded conductors of every other system on the planet via EGC. Whether its one-end grounded, center grounded, corner grounded, mid-point grounded, etc. doesn't make a difference to the other side of the transformer.


EGC on the secondary, that's bonded to the case, that's bonded to the X0, that goes back to the gear, that dead-ends there and has no galvanic connection back to the MV system....

Similarly, you may consider EGC on the secondary, that's bonded to the case, that's bonded to the earth, that's connected to all other grounded transformers on the planet... What difference does it make?


Define "everything" and clarify what you mean by how a corner grounded system "acts"...

Xptp and all that have responded, thank you for your patience and understanding.

What I mean by 'everything', is your unintentional fault now puts 2.3kV potential on all l-n loads upstream, there's disconnects, fire suppression controls, low voltage circuit. I'm not seeing how the 2.3kV interacts with the 480/277 bonds, other than since no current is flowing on the 2.3kV, it's just there.

If a second 2.3kV goes to ground, its p-p fault condition, what happens to that energy? Is th ocp fast enough that low voltage equipment won't see any surge?

 

[xptpcrewx]

Xptp and all that have responded, thank you for your patience and understanding.

What I mean by 'everything', is your unintentional fault now puts 2.3kV potential on all l-n loads upstream, there's disconnects, fire suppression controls, low voltage circuit. I'm not seeing how the 2.3kV interacts with the 480/277 bonds, other than since no current is flowing on the 2.3kV, it's just there.

It doesn't put that stress potential on L-N loads upstream. The 2.3-kV appears everywhere throughout the system and therefore there is no difference of potential. This is the one of the main concepts behind bonding/grounding - to anchor systems to the same potential. If everything rises or falls together, there is no issue (like a bird on a wire). The only thing getting stressed in this case is the MV insulation because now instead of seeing line-to-ground voltage of somewhere near 1.3-kV, it now sees the full 2.3-kV.

If a second 2.3kV goes to ground, its p-p fault condition, what happens to that energy? Is th ocp fast enough that low voltage equipment won't see any surge?

It blows-up and depending on the transformer connection type, the LV system may see the fault transformed/phase shifted a particular way. Whether something is fast enough depends on the settings you've put into the breaker. Note: Per the nameplate, this transformer is internally fused on the delta, so it could be a racing condition between a LV breaker and local fuse to clear the fault.