Single Line to Ground Short Circuit 1

[gvasiliou]

Hi guys,

I'm George, kind of newbie in this forum and i would like some of your help, if this ok with you.

The Project: To determine the bolted L-G SCC, in a generator parallel to the grid configuration.

The Problem:
My hand calculations can not verify the results of the SC software used (ETAP).
I'm trying to verify Bolted LG SCC using MVA method and even calculating impedances (z) in per unit system using symmetrical components analysis.
In all cases, i can verify the overall SCC (85,8 kA at 400Volts bus 3) but i'm not able to verify that SCC coming from Generator is 46 kA and SCC coming from the Grid is 40,5 kA (this numbers are given by ETAP Software).

Anybody that could "play" with my system to advise if he is able to "verify" ETAP results (46 and 40.5 kA respectivelly)?

I have upload a file showing the system single line drawing with ETAP Results, the tech data used by ETAP to perform calculations acc to IEC909, and my hand notes proving the total L-G SCC at bus 3 , equals to 85,8kA (close to 86,4 of ETAP).

Thanks in advance for your support in this ball breaking issue.

George V - Hellas 2012

 

[gvasiliou]

Anybody got an idea...?

 

[ijl]

Maybe. The calculation method of iec60909 (and probably 909) contains several correction factors. For example, if the fault is between a generator and the step-up transformer, then a special correction factor is used. Because of this, the comparison with hand calculations may not be quite straightforward. Try to switch off the factors in ETAP, if possible.

 

[gvasiliou]

Hi "ijl" , thanks for your reply.

Actually the only parameters that you can adjust within ETAP are:
(a) Voltage Factor (c). Inside ETAP (c) is considered to be zero since for LV side and 1.1 for HV side. In my hand calcs i use the same factors.
(b) There is a parameter called contribution level and is set to 1.
(c) There is a parameter called "include fault impedance Zf" which is not selected, thus ignored. (same in my hand calcs).
(d) Another parameter called "X/R used for Peak calculation" is set to "method c". Even if you select method a or b the results are the same since the results mentioned before are RMS results and not peak results (same on hand calcs).

PS1: IEC60909 and 909 as you mention is the same IEC.

PS2: Seems strange to me that every other ETAP result (i.e 3ph bolted SC at LV Side or 20kV side) can be verified (or come really close) by hand calcs.
But on LG SC on the LV Side the difference of hand calcs vs ETAP Results are huge (talking about partial currents - total current can be proved as you can see in my notes).

If you got any other idea .... let me know.
Thanks in advance.
George V.

 

[Kiribanda]

gvasiliou,
May I know what type of fault at Bus#3 you had simulated in your model in order get 40.5 & 46 kA contributions from Utility & Gen respectively?

 

[jghrist]

To get the contribution from each source requires calculating the current through each leg of the positive-, negative-, and zero-sequence networks. Then you combine the sequence currents of each leg using the "a" operator (120°) to get the phase currents.

Is this how you have done the symmetrical component analysis? I'm not proficient with the MVA method, but I doubt that it can be used to get the individual contributions for a line-to-ground fault.

I couldn't open the .rar file.

 

[GEORGE2311]

Your transformer has a D winding in low voltage as i see in one of your sketch so i will not expect to feed a ground fault. What are the trafo winding settings in ETAP?

G.F.

http://www.engineering-services.gr

 

[gvasiliou]

Hi guys,
Thanks for your answers.
Find some clarifications hereunder:
@Kiribanda : Single Line to Ground Fault.

@jghrist: My hand calcs are ignoring actually tha "a" factor and that could be why the difference is big . Just remember i have difference only in the partial amps. Total Amps of my hand calcs are really close to the ETAP Results.
PS : You need the free "winrar" software to open rar files. Highly recommended ;-)

@GEORGE2311: That's correct. My T/F has a D winding on the LV side, and as a result Zero Sequence of the T/F is not feeding my short circuit at the LV Side.
But Positive and Negative sequence of the T/F feeds the SC a lot.
On the other hand total Zero Sequence is not "zero" since my Generator is Star - Solid Grounded and feed the LV bus 3. So the total zero impedance of the system is actually equal to the Generator Zero Impedance (since t/f can not reach SC location). This "soup" comes to the result you can see in my files.

Regards

 

[gvasiliou]

For all those people who would like to help and give some feedback to this issue, i have reload the attached file in zip format instead of rar foramt of the initial post.

Any suggestions are welcome.

G.V

 

[gvasiliou]

Note for "jghrist" comment:

The fault simulated in ETAP refers to the first phase, since is the maximum fault current.
Fault current in phase B and C is minimum compared to the first phase.

According to the IEC 909, in first phase the fault is equal to the sum of the Pos.Seq. Fault + Neg.Seq Fault + Zero Seq. Fault.

"a operator" is not involved for this first phase.

"a" operator is actually involved if you try to calculate the currents in phase b and c (or L2 - L3 whatever you like).

So i don't think that the huge difference of my hand calcs VS ETAP results comes out from the missing "a operator" in my hand simplified calculations.

BR

 

[jghrist]

I haven't gone through all of your calculations, but I don't see where you have calculated the sequence currents through each leg of the top diagram in 3.8.2. This is necessary to get the fault current in each leg. Overall, I0=I1=I2 because they are in series, but this is not true in each leg.

 

[gvasiliou]

Hi jghrist,

The approach is a little bit different than standard calculation method.

Actually i'm using symmetrical components analysis in order to calculate total circuit impedance (let's call it "Ztotal") in the following steps:
First Calculate Pos.Sequence total impedance (Zpos)
Then Calculate Neg.Sequence total impedance (Zneg)
Finally Calculate Zero Sequence total impedance (Zzero)
(in my case this is only the zero sequence impedance of the generator since grid is isolated by the delta winding of the transformer in the low voltage side)

As a result, total impedance Ztotal equals to :
Zpos + Zneg + Zzero.
Now i can calculate the total current:
I(total) = I(1) = I(2) = I(0)

Since now the current of the whole circuit is calculated (using total impedance Ztot = Zpos + Zneg + Zzero) i can go back to each branch analysis (pos branch/neg. branch /zero branch) and apply this total current to each "leg" and split it proportionally (see the new attached page).

At the end of the page you can see the deviations from ETAP, calculating each leg current.

 

[jghrist]

You have to use complex impedances, not the magnitudes. The source and transformers have resistances.

 

[gvasiliou]

You are right about that.
On the other hand, all my searches about this story (before to post it) concluded to the fact that resistance components of sources / transformers is too small compared to the reactance components and it might be neglected.

It seems that neglecting resistance is OK when you want a "rough" estimation of the total fault current on a bus (proved by the fact that total fault current is not so far from ETAP results).

Maybe for more "accurate" calculations (like "leg" calculations), resistances has to come into play.

I was hopping to avoid this part...

 

[ijl]

Your splitting factor may not be correct. Shouldn't it be 2.84/(6.84+2.84) ?

 

[gvasiliou]

IJL it seems you are absolutely right.

I really thank you first for taking time to read my notes and secondly for your correction.

The correct split factor is 2.84/(6.82+2.84) = 2.84/ 9.68 = 0.2933

Applying this factor , then YES the hand result are REALLY CLOSE to the ETAP Results. (see attached corrected page).

Really thank you for the tip.


Regards