Modelling generator saturation in short circuit calcs ?

[rsherry]

The machine reactances used for calculating fault currents in Industrial systems are usually the saturated reactances. The UK standard (ER G74) states (para 9.1.3) that unsaturated machine reactances should be used for fault current calculations for EHV faults (i.e. on the 400 kV Transmission system). Does anyone have any comments on this ? Anyone know of a reference source to back this up ?

 

[dpc]

I have bumped into this question myself and found it a little hard to pin down in the normal reference books.

However, IEEE Standard 399, Power System Analysis (Brown Book), says that the rated-voltage (saturated) value should be used for short-circuit analysis. Also, IEEE Std 141 (Red Book) says that Xdv" should be used instead of Xdi", which I take to mean the rated-voltage measurement (which would be saturated) rather than the rated-current measurement, which would be unsaturated.

I think this is done because it is more conservative. All of these reactances are greatly simplfied representations of a complex electromagnetic system, but they have proven to be useful approximations over many years of use.

So.. I vote for saturated reactance values for short circuit calcs. Any other thoughts?

 

[Modula2]

At the high currents for short circuits, the machine reactances would be expected to be saturated.

 

[rsherry]

Thanks for the comments.
The reason given in the G74 standard for the use of unsaturated reactances is that machines (with typical generator transformers) would only see about a 50% terminal voltage change for a Grid fault.
The other argument I've heard is that the flux path involved for the subtransient and transient reactances would only saturate at high stator currents - so for a fault on the terminals there would be 4-5 times rated current (i.e. use saturated reactances for these), for a Grid fault there would be 2-2.5 times rated current and this may (?) not saturate these paths.
Given the variations with machine operating point etc I am certainly inclined to use the pessimistic values (saturated) but unfortunately that isn't in line with this standard.
Like dpc I haven't found much in the standard reference sources to clarify things.

 

[Alex68]

You are applying the UK standard ER G74 (1992), whose title is "Procedure to meet the requirements of the IEC909...", so it refers to the IEC909-1988. This standard is now substituted by the new version EN60909-0 issued in Dec. 2001. Thus the ERG74 is not valid now, I think!

The new version, IEC EN60909-0, states in the scope (1.1):
"This part is applicable ... in HV 3ph a.c. systems. Systems at highest voltages of 550kV and above WITH very long transmission lines need special consideration."
Excluding this particular case, this standard is valid.
Please note that some formulas are canghed from the old version, but you must always use the saturated reactances.

Two comments:
1- the "magic" voltage factor c, includes (and involves and solves) a lot of problems; in its definition, you read:" the... voltage factor c is necessary for various reason. These are: .... the subtransient behaviour of generators and motors"
2- the generator contribution to a fault in a HV system comes through a step up transformer; so you should consider the formulas referring to the "power stations"

In conclusion I think you must use the saturated values!

 

[jbartos]

Suggestion: Visit

http://standards.ieee.org/colorbooks/sampler/VioletBook.pdf

for:
Saturated machine reactance use:
1.2.15 direct axis substranient reactance: X”dv - (saturated, rated voltage) is the apparent reactance of the
stator winding at the instant short-circuit occurs with the machine at rated voltage, no load. This reactance
determines the current flow during the first few cycles after short-circuit.

1.2.18 direct axis transient reactance: X'dv - (saturated, rated voltage) is the apparent reactance of the
stator winding several cycles after initiation of the fault with the machine at rated voltage, no load. The
time period for which the reactance may be considered X'dv can be up to a half (1/2) second or longer,
depending upon the design of the machine and is determined by the machine direct-axis transient time

Unsaturated machine reactance use:
1.2.16 direct axis subtransient reactance: X”di - (unsaturated, rated current) is the reactance which is
determined from the ratio of an initial reduced voltage open circuit condition and the currents from a threephase fault at the machine terminals at rated frequency. The initial open circuit voltage is adjusted so that rated current is obtained. The impedance is determined from the currents during the first few cycles.


1.2.17 direct axis transient reactance: X'di - (unsaturated, rated current) is the reactance which is determined from the ratio of an initial reduced voltage open circuit condition and the currents from a threephase fault at the machine terminals at rated frequency. The initial open circuit voltage is adjusted so that rated current is obtained. The initial high decrement currents during the first few cycles are neglected.

 

[Alex68]

I'm sorry, jbartos, but I disagree with you.
We cannot confuse the IEEE calculation method with the IEC one. They are based on different approaches.
Rsherry is applying the G74, which refers to the IEC method.

The concept reported in the Violet Book are correct but the IEC standard requires the saturated values in its formulas, which contain some proper parameter to correct their effects. If don't respect them, you will obtain wrong results (as I experienced with dramatical consequences).

 

[Modula2]

So, what is the minimum fault contribution from a generator, such as one not directly at the fault, maybe even away from it. What do the standards say about that? At what point away from the fault do we switch from sat to unsat. For relaying, it's good to know.

 

[Alex68]

Modula2
Are you referring to the IEC60909-0?
If so, please refer to point 2.5 in which you find:
"when calculating minimum s.c. currents, it is necessary to introduce the following conditions:
- voltage factor Cmin...
- ... system configuration...
- motors neglected
- lines resistances... at higher temperature..."

No mention about sat. or unsat. reactances or distance!
Please remember that is a simplified method bringing very realistic but approximated results.

About the minimum currents for the protection setting, my suggestion is to consider the minimum currents in 2ph (and not 3ph) short circuits.