Upgrading Utility Distribution Transformer with an NGR

[rockman7892]

A utility customer is considering installing an neutral ground reactor on one of their distribution transformers (12.47kV).

I'm familiar with NGR in several other applications but have not seen many of these installed on distribution transformers. Is it common for utility's to have NGR's on their distribution transformers to limit fault currents or help with coordination?

I know in most industrial facilities the NEC does not permit NGR's to be installed on circuits with L-N loads, and I'm assuming with a utility distribution circuit there would be a similar requirement to prevent a voltage potential between N-G on the system during normal conditions with neutral current flowing through NGR?

Is approach to sizing these NGR's similar to other applications where you evaluate simultaneous objectives of keeping fault current to a minimum on system, while at the same time allowing enough fault current to make sure relays operate in a coordinated fashion? I assume this would be modeling the entire distribution circuit to see what faults on remote end of distribution system appear as for relaying?

If a NGR is being placed on an existing OH distribution line wouldn't the PT's, arrestors, CPT's and other equipment need to be evaluated for adequate ratings for L-G connections during faults when L-G voltage is elevated to L-L (IE proper bushing, winding ratings, etc..)

 

[waross]

As I understand the rules, fault currents are based on line to line faults, not line to ground or line to neutral faults.
If they are impedance grounding distribution circuits I hope that the services do not include a neutral conductor.
The last time I saw impedance grounding it was on the power transformers of a world class mine mill.
There were voltage relays across the impedances.
In the event that a piece of equipment developed a serious ground fault, the impedance ground avoided a general power outage and allowed time to isolate and remove the faulted equipment.
The voltage relays produced an alarm in the event of fault current through the impedance.
These were on a 13,000 Volt secondary system feeding 13,000;480 Volt unit subs..

 

[rockman7892]

Is there an advantage to having a neutral reactor as opposed to a neutral resistor on a distribution circuit?

From what I've seen this is commonly done with generators where the Z0 is much less than Z1 causing zero sequence current to be much higher than positive sequence current and this the generator may not be properly braced for zero sequence. In these instances is appears that the generator can still connect L-N loads? Does a neutral grounding reactor still allow the connection of L-N loads in contrast with a grounding resistor?

It appears that the sweet spot for reactors is limiting the gf current to no less than 25% with 25%-60% being the target range.

 

[waross]

Is there an advantage to having a neutral reactor as opposed to a neutral resisto

Less heat to dissipate.

 

[Kiribanda]

Typically, not like phase faults, all line to ground fault currents on any distribution system are limited by the fault resistance. Therefore when you introduce an NGR then those GF currents are further limited and might go undetected. Therefore, unless a GF monitoring system is introduced it is difficult to sense those already limited GF currents. Hence I donot think there is any advantage in using NG reactor/ resister unless it is proven by system sc studies that due to system growth the equipment ratings are not sufficient now. Therefore, current limitation is justified. Also, when a reactor is introduced there is a possibility for resonance between the reactor & 12.47kV cables (if any) capacitance currents. All these will be revealed by system studies.

 

[waross]

Hence I donot think there is any advantage in using NG reactor/ resister unless it is proven by system sc studies that due to system growth the equipment ratings are not sufficient

As I understand the rules, <system> fault currents are based on line to line faults, not line to ground or line to neutral faults.

 

[Kiribanda]

System growth can happen by adding new generators at the upstream. As result the ultimate ground fault level (also 3-Ph) given today may be different after 10yrs time depending on the system growth.

 

[waross]

System growth can happen by adding new generators at the upstream. As result the ultimate ground fault level (also 3-Ph) given today may be different after 10yrs time depending on the system growth.

When considering fault current levels, line to line faults are considered, not line to ground faults.
If system upgrades increase the level of available line to line fault currents, then reducing line to ground fault currents does nothing to help.

 

[wcaseyharman]

My utility often installs neutral reactors in distribution substations, 0.5 ohms typically. I remember hearing it was to limit the LG fault current to prevent damage to concentric neutrals on cables during LG faults, but I need to verify that.

 

[waross]

That sounds very reasonable, Casey.

 

[genhead]

Rockman7892 asked - Does a neutral grounding reactor still allow the connection of L-N loads in contrast with a grounding resistor?
Does anybody have any comments?

 

[waross]

NGRs are more at higher voltages where there are no line to neutral loads.
On distribution circuits with transformers connected line to neutral, there is generally distributed grounding of the neutral and grounding of the neutral at each transformer. An NGR would not work well in that situation. Much of the fault current will be arriving via the neutral conductor.
Only consider an NGR where there are no line to neutral loads.

 

[genhead]

I knew that to be so with resistors. I understand that line-to-ground loads/faults will cause rising line-to-ground voltages on the unloaded/faulted phases. Is the effect any different with a reactor as opposed to a resistor?

 

[waross]

For a given current a reactor generates less heat than a resistor.

 

[Amp Dylan]

Not typical for 12.47kV distro unless you’ve got no L-N loads and a specific need to limit LG fault current. Gotta model the whole system and check equipment ratings especially PTs and arresters. Usually better suited to industrial setups or higher voltages.

 

[dpc]

Why do they think they need a grounding reactor? It's uncommon on utility distribution feeders, except for short underground feeders with high fault current. With overhead lines, more ground fault current is usually better. I've personally never seen a reactor on a standard utility distribution feeder.

 

[rockman7892]

The distribution circuits in this application are fed from a sub what has a 138kV wye - 13.8kV delta transformer with a grounding transformer connected to delta side of transformer. So not a conventional wye grounded substation but rather a 13.8kV delta with grounding transformer.

NGR was proposed to reduce fault magnitude and nuisance tripping for faults close to substation.

With a grounding transformer can their be L-N connected loads on the system? If an NGR is added to grounding transformer do the same considerations for L-N loads and equipment having increased ratings for L-L voltage elevation during faults apply?

 

[rockman7892]

I knew that to be so with resistors. I understand that line-to-ground loads/faults will cause rising line-to-ground voltages on the unloaded/faulted phases. Is the effect any different with a reactor as opposed to a resistor?

My understanding has always been that L-N loads are prohibited due to the fact that normal neutral current will flow through resistor and cause voltage drop across resistor thus creating a dangerous voltage potential between neutral and ground. That was always my understanding in the industrial NEC world but imagine same danger potential exists on distribution. I'm also not sure if that changes with a reactor or same voltage potential hazard would apply.

 

[waross]

The grounding transformer changes the considerations.
If a grounding transformer is used to support line to neutral loads on a delta connection, it must be sized to carry not only greatest possible neutral current (single phase event) but also must be able to withstand possible fault current levels.
There should not be any impedance grounding used wiyh line to neutral loads.
If a grounding transformer is used to support detection and reaction to line to ground faults then a neutral grounding impedance limits the fault to ground current and allows a much smaller grounding transformer to be used.
In the case of a large system with relatively large charging and leakage currents, the use of a reactor rather than a resistor will result in less heat being developed.
This heat may have non-electrical considerations, such as possible damage to a nearby concrete wall by elevated temperatures.
I don't know the exact details as it was not part of my assignment, but I was on a project where other engineers mentioned that they were concerned with possible damage to concrete by energy dissipating resistors.
This was on an electric transit system. At times when regeneration exceeded draw, excess energy was dissipated in resistor banks. Reactors would not be suitable for the dissipation of energy.
The point of this anecdote is not the choice between resistor and reactor but the issue of heat rejection from resistor banks.