Zero sequence return current share 2

[Power0020]

For the network hereunder, a ground fault at "F" will return to the source transformer neutral path indicated as path #"1". The equivalent zero sequence network indicate that all this current is going back to the Delta-Wye transformer between bus (A) and (B) as shown in path #"2".

It may not appear in the zero sequence network that the fault current is shared between both transformer neutrals. This makes it misleading in grounding design?

https://files.engineering.com/getfi...c22-036243d84e82&file=IMG_20210128_181542.jpg

any clue?

 

[jghrist]

Some of the return current will flow in the neutral and/or ground wire between F and E, and some will flow through the earth from F to E. Some will flow through the earth between F and B. Some of the current flowing from F to E will flow to B through the neutral and/or ground wire between C and B. Some will flow to the wye-delta ground source between C and D. Some will flow through the earth between C and B.

Your typical zero-sequence network diagram does not distinguish between these paths. The split depends on the mutual impedances from the faulted phase wires to earth, from the phase wires to the neutral and/or ground wires, and from the neutral/ground wires to earth. All very complicated. It can be calculated with sophisticate software like SGSYS (Substation Grounding SYStem Analysis Program) or CDEGS (Current Distribution, Electromagnetic Interference, Grounding and Soil Structure Analysis). IEEE Std 80 has some tables that might help estimating the split.

 

[waross]

As well, there will be a heavy circulating current in the delta of the wye/delta transformer and there will be a back-feed from the wye/delta transformer to the fault, increasing the fault current.

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

 

[EddyWirbelstrom]

See attached files for SLD and report showing branch currents for earth fault at BUS_E.
Send equipment data if you require actual currents.
The Comprehensive method of PTW32 Ver 9011 was used.

It appears only one file can be uploaded.
I will send the SLD in a separate post.

 

[EddyWirbelstrom]

Attached is SLD

 

[Power0020]

Thanks EddyWirbelstorm, much appreciated.

Can you please check the neutral currents within this network, no equipment data as this is a typical network just to understand the split, assume that there is no any ground wires for TLs.

 

[EddyWirbelstrom]

My example does not consider earth fault current flowing in the mass of earth, it only considers earth fault current flowing in the conductors.
In your case - with no earth/cable screen path - the return earth fault current would split between the neutral and mass of earth.
The split factor ( Sf ) is the ratio of earth fault current flowing in the earth ( Ig ) to the total earth fault current ( 3 Io ) .
Sf = Ig / 3 Io

The split factor depends upon:
̵ Earth grid resistance ( Rg ),
̵ Conductive split factor – where Ig = ( Neutral Z / ( Neutral Z + Rg ) ) 3 Io
̵ Mutual coupling between phase conductor and neutral conductor and earth.

Please refer to Eng-Tips thead203435 dated 27/11/07

 

[Power0020]

Thanks Eddy, I am aware of the fault current split factor between grid and earth wires.

My questions is about the fault current share between transformer neutrals upstream to the fault. e.g. 40% of the fault current returns to XFM_Gen 01 neutral and 60% to XFM_Yy neutral, if any?

 

[EddyWirbelstrom]

The currents flowing in the transformer neutrals are shown highlighted in the attached file.
These currents will split between the mass of earth and the earth/screen/neutral currents.

 

[waross]

Current splitting:
This is a factor on long runs.
For short runs, with a typical 5 Ohm resistance at the grounding electrodes the current through the ground path is often negligible.
If you are calculating actual currents, you may be neglecting a contribution of about 175 Amp from the wye/delta transformer.

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

 

[Power0020]

Much appreciated Eddy, this says a lot.

I am a bit astonished to see a circulating current at the delta of XFM_YNd11 due to the SLG fault downstream XFM_Yy ( I guess it will be about 757 A instead of shown 1360 A). However this is seems reasonable for the magnetic balance of the transformer. Opening this neutral connection will simply mean that all the 22 kV zero sequence current will feedback into XFM_Gen01 neutral.

With the SLG fault current at the 22 bus of about 1300 A, the neutral current for faults at 415 V bus is about 15% of that. I guess the simulation tells why it is always good to use a delta winding in the direction of power flow, as the ground fault protection for the XFM_Gen01 transformer will need to be de-sensitized to faults downstream XFM_Yy.

 

[EddyWirbelstrom]

Waross
Thank you for your helpful post.
Can you please expand on your comment ‘175 Amp from the wye/delta transformer’. Is it the XFM_YNd11 transformer ?

Power0020
You are correct, the current circulating in the delta of the transformer XFM_YNd11 should be 1368/sqrt(3) = 790A. Attached is the corrected SLD which also includes current in the delta winding of transformer XFM_Gen 01..
Can you please expand on your comment ‘the ground fault protection for the XFM_Gen01 transformer will need to be de-sensitized to faults downstream XFM_Yy’.
Do you wish me to run a short-circuit study with current distribution for an SLG at 22kV BUS_C ?
SLG at BUS_C is 693A.

 

[Power0020]

If there is a fault downstream, about 15% is flowing into the transformer neutral upstream as the example shown. If the ground fault protection is set with a setting less than 15% of the rated SLG it will falsely trip and loose selectivity. So it has to have a setting higher than 15% of the nominal SLG of the transformer secondary fault level, which on other hand may limit the protection operation if an impedance fault takes place.

Please proceed with a fault distribution on the 22 kV bus, I'd expect a very small contribution from the LV bus.

 

[EddyWirbelstrom]

I have added an SLD showing zero-sequence current distribution for a SLG at BUS_C to Rev C of attached file.