Transmission impedence vs. rated MVA capability

[ehanes75]

Greetings,

I am working on verifing impedence data for a large number of transmission transformers. The data is being used in a GE PSLF powerflow model of a large transmission nework.
I would like to develop quick way to flag any bad data. In the interest of time, I can't go pull the test report info for 500+ transformers. Besides, some of these units are 80+ years old and no test report data exists.

The transformers vary in size from 1 MVA generator step up units to large 1000+ MVA 500/230 kV units.

What are reasonable values for R and X?
How does R and X vary with transformer size?
What about X/R ratio?
I imagine that turns ratio would also affect these values.


I look forward to your response!

Thanks

 

[davidbeach]

They can vary tremendously. Try Conrad St. Pierre's A Practical Guide to Short-Circuit Calculations.

 

[cuky2000]

1) For X/R:

http://cuky2000.250free.com/XFMR_XoverR.jpg?view_image=1

2) Standard transformer impedance are shown in the following link.

http://cuky2000.250free.com/Transformer Impedance.jpg

The data in this table could be adjusted if transformer with load tap changer (LTC), different than selft coolling rating or rating temperature as follow:
1- For LTC equipment add 0.5% to the value listed above.
2- To obtain Std impedance for FOA or FOW multiply by 1.667 the value listed above.
3- To Obtain std impedance at 65 oC, multiply by 1.12 the listed value above.

 

[tlr7616]

Your best bet is to us what you have--you say you have a large case in hand, presumably it is a PSLF *.sav or *.epc case.

Open it in PSLF, go to the TRANS table, and start looking at the data. I would suggest exporting that data to Excel so you can perform some calculations, sorting, filtering, etc on the data.

Anyway, you should sort the transformer data by transformation voltage, and then by MVA range. You will see that the X falls into a definite range of values, the statistical variations should become apparent.

Below are some thoughts on your specific questions:

1. Reasonable values for R and X?

Some modelers will use an ideal transformer model, meaning R=0. Not totally correct but frequently happens--note this screws up X/R calculations in a big way! X should be positive, note that three winding transformers will have an element with -X and -R at times, the math works out but it is in load flow cases frequently and it is confusing. R's typically vary from 0.0001 (i.e. lossless) to perhaps 4.0 or so. X varies from 0.0001 to perhaps 10. Note that this data is across a range of transformation volage and MVA sizes. The proper way to evaluate this stuff is to look at the charts noted previously and use the percent impedance on the transformer MVA base.


2. How does R and X vary with transformer size?

As the MVA size goes up, R tends to go down. Note that like all things there are exceptions, old transformers might be more lossey than newer ones. However, if you compare a 100 MVA bank against a 1300 MVA you will see that the per unit R for the large bank is very small (i.e. ~0.0001). The same behavior holds true for X a well.


3. Range of X/R ratio's?

First off, you need a non-zero R for this one...
Second, this is highly variable, the X/R ratio works fairly well for transmission lines, not so good for transformers. I have seen transformer X/R's in a case range from less than 1 to 3000. If I had to give an average I would say perhaps 50--it is all over the place.

Again, I would suggest to you that if you have access to a large load flow case that you can get a good sense of what the statisical norms are, but do it by transformation voltage and MVA rating range.

You should see a strong trend line for X and R in Excel when you do that. Recognize that you will see many outliers. Also, realize that these load flow cases are full of "modeling slop" and as such you have to be aware of what data might influence your study and what does not--it is an issue of sensitivity.

Oh yes, off nominal turns ratios and tapping has an influence on this, but it is relatively minor and should be handled by the impedance correction table (i.e.tztab).

Let me know if this was helpful.....