Three-winding transformer and charging current on resistance grounded secondary

[pejap]

Hi,

I need to add a NGR (neutral grounding resistor) to a Customer specified three-winding transformer with following parameters:
Primary: 46kV, wye-connected with solidly grounded neutral. 46kV system is solidly grounded.
Secondary: 4.16kV, wye-connected. To add NGR on this winding.
Buried delta tertiary (or stabilizing winding): 4.16kV delta connected, corner grounded
Power: 6 MVA
Vector group: YNyn0(d11)

This transformer is used to feed industrial loads and resistance system is also a requirement. NGR will be rated for 10 seconds and the system will trip on the first ground fault.

When it comes to resistance grounded systems in industrial distribution , I am used to working with two-winding transformers containing a delta in one of the windings, which blocks zero sequence current flow from primary system to secondary system and creates “ground fault islands”. In these systems, I only use secondary capacitances to determine the NGR rating.

In this case, however, I am not so confident and I need a validation. I believe that primary charging current (e.g. resulting from 46kV system stray capacitances) will not enter the secondary system during single-line-to-ground fault on secondary (4.16kV) side. I believe that I can calculate secondary charging current by using secondary system capacitances only.

This is what I think happens and I would like your validation or correction:

For this three-winding transformer, the zero sequence model of the transformer connects primary to secondary sides of the system with a T-model where a tertiary zero sequence Z0 impedance is connected to ground.

When I draw a sequence diagram for single-line-to-ground fault on 4.16kV wye (NGR grounded) side of the transformer, the zero sequence charging current from primary side of three-winding transformer would take a path of least impedance and return to primary side of the transformer (e.g. to zero sequence capacitive reactance) via the small tertiary Z0 impedance rather than take a path through secondary and large NGR impedance. In this case, my NGR is sized at 160 ohms and the zero sequence current would see 3*160 = 480 ohms. This NGR impedance is much larger than tertiary Z0 impedance.

1. Am I correct in my assessment or is there another explanation for what happens here?

2. Am I correct in saying that to calculate NGR rating for connection to secondary 4.16kV wye system, I can consider only secondary (4.16kV system) charging currents?

3. For my own curiosity, if this transformer was a two-winding transformer, wye-to-wye where primary was solidly grounded and secondary needs a NGR, then how do you calculate charging current on secondary side of the transformer?
In other words, is there an influence of primary charging current onto secondary charging current?

Thank you

 

[jghrist]

Typically, low resistance grounding is used for medium voltage systems, not high resistance grounding. Fault current would be limited to 100 - 400 A and charging current is not a factor. See

https://megaresistors.com/products/...ystem Neutral Grounding Resistor,2400 / 400 =

 

[pejap]

Thank you for responding. Different industries take different approaches depending on experience.
Your typical suggested values of 100-400A is completely reasonable in some industries where I worked. Lots of literature (IEEE papers and books) recommend 400A value (as a seemingly standard practice), to limit transient overvoltage to 250% peak line-to-ground and to allow enough current to flow so that ground fault relays (usually electromechanical) operate reliably and with coordination to downstream ground fault relaying.

On the other hand, a modified high resistance grounding system can be used for MV systems where the intention it to limit equipment damage due to ground faults. To my knowledge this method was described first in 1991 IEEE paper “Selection of system neutral grounding resistor and ground fault protection for industrial power systems”. This is only possible due to advances in ground fault relays that can reliably detect low ground fault currents. In my experience, a 10A or 15A NGR is sufficient on most 4.16kV systems and 20-50A was sufficient on most 13.8kV system. In all those cases we used a two-winding transformer with one winding as a delta. In each case it is necessary to calculate the charging current in order to determine the rating of the NGR. And in each case we trip on a first detected ground fault.

On this system I am intending to use a modified high resistance grounding system to reduce burning damage to equipment on single line to ground faults. This is why I am asking about charging currents on this three-winding transformer.

 

[waross]

Your 4.16 kV system will have an impedance to ground on each 2.4kV phase to ground.
The resulting leakage currents may act as a wye connected three phase load and be seen by the protection CTs as a three phase load.
In the event of a line to ground fault, charging current will flow,limited by the line to ground impedance plus the impedance of the NGR.
But, in the event of a line to ground fault, a line to ground current limited only by the NGR will flow.
I don't see a correlation between charging current and line to ground fault current.
That said, as the physical size of a system increases, so often, does the charging current.
I suggest that basing protection settings on charging current is a good way to avoid nuisance trips.

The delta:
A phase loss in a wye/wye system may cause the loss of charging current on one phase.
This in turn may cause a ground fault trip on phase loss.
Setting the trip level above the level of charging current will avoid ground fault trips on phase loss.

You don't have a wye/wye transformer, you have a wye/delta/wye transformer.
In the event of a phase loss on a wye/delta or wye/delta/wye transformer, the delta winding will rob power from the healthy phases to supply the missing phase.
In the event of a bolted line to ground fault, the delta contribution may increase the fault current above the level of a bolted three phase fault.
In the event of a primary phase loss and a light load on the 4.16 kV system, the delta contribution may hold the secondary at very close to full voltage. (within a few percent of full voltage.)

What does this mean to you?
Much less chance of a nuisance ground trip during a phase loss event.
A setting base on charging current is still a good idea.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

The wye/delta:
Why does the delta replace a lost phase?
Consider two 2400 V to 240 V transformers connected in open delta.
There will be 240 Volts available to supply 240 Volt loads across the open side of the delta.
Suppose we connect a 240 V to 2400 V transformer across the open delta.
That transformer will develop 2400 Volts on the high voltage winding.
We can supply loads with that 2400 Volts.
But wait. A three phase bank with a missing primary phase has a 240/2400 Volt transformer already connected across what becomes the open delta when one phase is lost.
You will see full primary voltage developed and back fed into the primary circuit that will feed any other loads between the transformer bank and the point where the circuit went open.
I spent over 15 years in Why/Delta land and have seen this too many times to count.
In a wye/delta/wye transformer, the secondary will also be energized by the action of the delta.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[dpc]

Determination of the effective charging capacitance is always difficult, more so with three-winding transformers. Core vs Shell construction will have an impact as well. I would inquire with the transformer manufacturer to get their best approximations. The capacitance values for the rest of the system should be similar to more conventional systems.

The hybrid grounding system you reference was intended for improved protection of industrial generators against damage from internal ground faults, IIRC.