Generators + Zig-Zag Tx: What to do with the neutrals? 2

[mbk2k3]

Hey folks,

I have 3x 13.8kV generators connected in parallel to a 13.8kV bus with a zig-zag transformer.
I'm trying to see what my options are here for the generator neutrals. See attached PDF for a sketch of what I'm trying to get at.
The way I see it I have 3 options: OPTION 1 - transformer in neutral, OPTION 2 - transformer with resistor, OPTION 3 - full blown NGR
[ol 1]
[li]What do I do with the generator neutrals to achieve quick tripping + selectivity?

​

e.g. If G2 has an internal ground fault, I want the ground fault current to be limited and only G2 to trip.

​

e.g. If the 13.8kV bus has a fault, I want all 3 generators to trip offline.
[/li]

​

[li]NGRs (OPTION 3) at 13.8kV becomes big and expensive. Can I avoid NGRs and stick to a single phase PT/grounding transformer with a resistor connected (OPTION 2) to limit the fault current?[/li]

​

[li]OPTION 1 would result in transient overvoltage conditions + ungrounded arcing ground fault condition after the generator breaker opens, no?[/li]
[/ol]

 

[RRaghunath]

You have a bus connected zigzag earthing transformer that provides earth reference and earth faut current path for 1-phase to ground faults in 13.8kV system. There is no need to earth the generator or transformer (if you have grid connected transformer) neutral any more. Just keep them all isolated.
To detect earth faults in individual generators, have directional earth fault protection on generator line side, looking towards the generator. Same goes with the transformer 13.8kV side earthfaults as well.
To set the earth fault protection properly, you need zero sequence impedance of zigzag transformer and details of NGR if any connected to its neutral.

 

[mbk2k3]

Thanks for the input.

From everything I've read, isn't it advised against running generators with open neutrals? Transient overvoltages if there are reoccuring arcing faults,etc.

 

[xnuke]

You're supposed to minimize the time the generator is running disconnected from the grounded bus. Get your machine up, synchronize it, and connect it quickly to the grounded bus. If you run it with the breaker open for a long period of time, your concerns might become valid, but nobody does that.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[mbk2k3]

@xnuke: interesting.

is this configuration really that common? I've only seen NGRs in the gen neutrals.
during commissioning/start-up/testing we do idle the engine for extended periods before closing the generator breaker. how does that play into this? is that considered a 'long period of time'?

 

[dpc]

It's fairly common to have a shared NGR for multiple generators on a bus. Not as common in the US. In this case, I'm assuming a ground reference is needed even if all generators are offline. I'd leave the generator neutrals ungrounded as others have suggested. As soon as you close the generator breaker, it is grounded. If you're concerned, you can use a broken delta PT arrangement and overvoltage element to detect ground faults in the generator prior to synchronizing.

 

[mbk2k3]

@dpc: thanks for the response. Does it have to be a broken delta PT arrangement? Or can i use a single PT (13.8kV to 120V) like what I show in OPTION 1 in my attachment on the first post.

 

[RRaghunath]

You can use single PT connected between generator neutral and earth for detecting generator earth faults when the generator is not yet connected to the busbars.
Since you already have a three single phase PTs at the incomer to the busbars, you can also use the (derived) residual over voltage function in the directional earth fault protection (59N) at the incomer. This way,you won't be needing an additional single phase PT.

 

[waross]

during commissioning/start-up/testing we do idle the engine for extended periods before closing the generator breaker.

No problem. If you are idling at low revs, the UFRO will reduce the voltage to keep the V/Hz ratio constant.
If you are idling at full speed and voltage, commissioning/start-up/testing is a good time to discover any internal issues.
In the days before UFRO became a standard feature, idling a set would destroy the AVR in much less than a minute.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[mbk2k3]

hey guys, i wanted to follow up and close the loop on this topic a bit:

I wanted to know how this arrangement (neutrals floating / open when the generator breaker is open) holds up when it comes to:
- Ground faults (single fault = no current, 2nd fault = phase-to-phase, high currents?), are Stamford/Cummins/AVK/similar alternators braced adequately for these types of faults?
- High transient voltages, how high of transients are talking? Are the winding insulations rated for a certain threshold? What is that threshold?
- Etc. any other concerns

NGR / neutral grounding resistor/reactor manufacturers are adamant that NGRs are necessary in the neutral, but it seems more of a recommendation than a requirement. Or am I misunderstanding them....

 

[dpc]

At higher voltages the line-to-ground fault current can exceed the three-phase fault current at the generator terminals. Generators are often braced only for three-phase faults so it is recommended to limit the line-to-ground fault current using a neutral to ground impedance of some kind. So the need depends on the design and bracing of the generator. But equally important is limiting the damage to the generator core from an internal fault. Since most faults at least start as ground faults, limiting ground fault current reduces machine damage from these faults.

IEEE guidelines recommend resistance grounding of all medium-voltage generators.

Running generators ungrounded for short periods of time prior to synchronizing should not be a major concern, at least in my opinion. Generator should be provided with surge protection at the generator in any event.

Continuous ungrounded operation increases the risk of voltage transients.

 

[mbk2k3]

@dpc, thanks for the response.

your response is pretty typical what i've been seeing.

everyone says "running ungrounded for short periods is ok" but then they also say "limit the damage to the generator core from internal fault".
these almost sound contradictory. if there is an internal fault and the generator breaker opens and I don't have resistance grounding on the neutral, i'm not limiting the damage in any way am i?

I'll be getting in touch with some manufacturers, but what am i looking for when it comes to "design and bracing of the generator" to see whether they can run ungrounded?

 

[dpc]

Once the generator breaker trips and the field trips, the fault current will decay based on the generator time constants. The bracing is a separate issue and is only applicable for faults external to the generator.

Generators spend 99% of their energized life connected to something. So we're talking about increased risk over a short time period. For complete protection, you can ground each generator individually and that is what I would do if starting from scratch. But for an existing system such as this, I don't the investment to modify it can be justified.

 

[mbk2k3]

@DPC, I shouldve stated earlier: this isn't an existing installation. its a new installation.

There are just logistical/physical constraints + financial reasons for pushing for the NGRs to be removed if we can still operate the system safely.