Transformer Impedance?

[mivey]

Perhaps- but why not 10% or 12.5%? I've seen 14.5% Z specd to control fault current on a 13.8kv bus. Then again 8% Z is better then 2% Z so you might be right.

Higher %Z yields higher losses and worse voltage regulation. Lower %Z can yield fault currents too high.

For substations, we traditionally try to limit fault current to 10 kA and this is sometimes achieved by requesting a higher impedance design. You have to consider the lowside equipment ratings and this used to be 8-10 kA or less so you would not want to expose these devices to higher fault currents. If you have higher rated equipment you can lower the impedance accordingly, at least to the transformer design limit.

 

[gar]

190302-0859 EST

mbrooke:

The flux in a core is proportional to the volt-time integral of the applied voltage. The integral of a sine wave is proportional to the cosine of a sine wave. This means that the integral of a sine wave becomes a maximum or minimum at the zero crossing of the sine wave. Integration is a means of measuring the area under the curve being integrated. If you integrate a sine wave starting at t=0, where t=0 is a positive zero crossing, then the integral keeps growing positive as you move to the next zero crossing reaching its maximum value at that negative zero crossing. Then on the negative side of the input sine wave it is subtracting. Its most negative value is at the next zero crossing which is at a positive slope.

This integration is why you have the volt/Hz constant.

How much you can over-voltage a core is a function of the shape of the magnetization curve, soft or square, and tolerable over power dissipation. Many factors here. Lifetime is modified.

.

 

[gar]

190302-1455 EST

To give you some idea what integration is all about consider 1/4 of a full cycle sine wave, 0 to Pi/2.

Integration is finding the area under a curve. Break the 1/4 sine wave into 9 rectangular pulses starting at 5 degrees and spaced 10 degrees apart. The sum of these small rectangular areas is an approximation of the area under the 1/4 sine area.

Angle ... Sine

05 ....... 0.0872
15 ....... 0.2588
25 ....... 0.4226
35 ....... 0.5736
45 ....... 0.7071
55 ....... 0.8192
65 ....... 0.9063
75 ....... 0.9659
85 ....... 0.9962

Sum .... 5.7369
Ave ..... 5.7369/9 = 0.6374 as compared to 2/. Pi from calculus = 0.6366
For an approximation this is a good coorelation.

.

 

[mbrooke]

Higher %Z yields higher losses and worse voltage regulation. Lower %Z can yield fault currents too high.

For substations, we traditionally try to limit fault current to 10 kA and this is sometimes achieved by requesting a higher impedance design. You have to consider the lowside equipment ratings and this used to be 8-10 kA or less so you would not want to expose these devices to higher fault currents. If you have higher rated equipment you can lower the impedance accordingly, at least to the transformer design limit.

How much worse is voltage regulation at 8% to say 12.5% from no load to full load?


And yup- fault current is a major driver, even in cases over 10ka.

 

[mbrooke]

If you look at ANSI/IEEE C57.12.10 it shows that, for 450 BIL, the impedance shall be 8.5% (without LTC) and 9.0% (with LTC) by default unless the user specifies another value. Additionally, the manufacturer must agree and the user should perform a system study to determine the proper value.

Why 8.5-9.0%. by default? Not exactly sure as the standard does not provide insight for the requirement. My guess would be: A performance acceptable, economical and practical design impedance that can be achieved for that BIL.

Question- do you have a list of the BIL vs impedance? Higher BIL increases the insulation thickness and thus lowers the flux density?

 

[mbrooke]

190302-0859 EST

mbrooke:

The flux in a core is proportional to the volt-time integral of the applied voltage. The integral of a sine wave is proportional to the cosine of a sine wave. This means that the integral of a sine wave becomes a maximum or minimum at the zero crossing of the sine wave. Integration is a means of measuring the area under the curve being integrated. If you integrate a sine wave starting at t=0, where t=0 is a positive zero crossing, then the integral keeps growing positive as you move to the next zero crossing reaching its maximum value at that negative zero crossing. Then on the negative side of the input sine wave it is subtracting. Its most negative value is at the next zero crossing which is at a positive slope.

This integration is why you have the volt/Hz constant.

How much you can over-voltage a core is a function of the shape of the magnetization curve, soft or square, and tolerable over power dissipation. Many factors here. Lifetime is modified.

.

Thank you.


Lifetime modified? :?

 

[xptpcrewx]

How much worse is voltage regulation at 8% to say 12.5% from no load to full load?

Exactly that. 8% to 12.5%.

 

[xptpcrewx]

Question- do you have a list of the BIL vs impedance? Higher BIL increases the insulation thickness and thus lowers the flux density?

I wouldn’t really think about it in terms of reducing flux density because that would be compensated for in the design. It’s more about the having more leakage flux. Flux that doesn’t contribute to core flux but still drops a little bit of voltage.

 

[mbrooke]

Exactly that. 8% to 12.5%.

Your kidding, right? If so thats just way to cool

 

[xptpcrewx]

Your kidding, right? If so thats just way to cool

Not kidding. %Z and %V are equivalent in per-unit.

 

[mbrooke]

I wouldn’t really think about it in terms of reducing flux density because that would be compensated for in the design. It’s more about the having more leakage flux. Flux that doesn’t contribute to core flux but still drops a little bit of voltage.

Does their rating change with MVA rating? Or its independent regardless?

 

[mbrooke]

Not kidding. %Z and %V are equivalent in per-unit.

So {(V no load - V full load) / V full load} x 100 = % voltage regulation?

 

[xptpcrewx]

Does their rating change with MVA rating? Or its independent regardless?

This table is for liquid-immersed power transformers only. MVA is irrelevant as far as ANSI/IEEE C57.12.10 is concerned; but may be different for other transformer types. I encourage you to refer to the other ANSI/IEEE C57.12.xx standards.

I see you are an EE. You should have enough background to take it from here....

 

[xptpcrewx]

So {(V no load - V full load) / V full load} x 100 = % voltage regulation?

There are two formulas for voltage regulation, voltage regulation up and voltage regulation down. This is the formula for voltage regulation up.

As an EE you should get in the habit of doing your research before you ask basic questions.

 

[mbrooke]

This table is for liquid-immersed power transformers only. MVA is irrelevant as far as ANSI/IEEE C57.12.10 is concerned; but may be different for other transformer types. I encourage you to refer to the other ANSI/IEEE C57.12.xx standards.

I see you are an EE. You should have enough background to take it from here....

Very good- I can do the rest from here.

There are two formulas for voltage regulation, voltage regulation up and voltage regulation down. This is the formula for voltage regulation up.

As an EE you should get in the habit of doing your research before you ask basic questions.

And, any equation like that can be applied "backward" :thumbsup: All I need to know is if my thinking is on the right track. Still in shock its that easy lol.

 

[xptpcrewx]

Very good- I can do the rest from here.



And, any equation like that can be applied "backward" :thumbsup: All I need to know is if my thinking is on the right track. Still in shock its that easy lol.

Sounds good. Any power engineering handbook/textbook will cover most of your questions.

 

[mbrooke]

Sounds good. Any power engineering handbook/textbook will cover most of your questions.

Well, only to a degree.

 

[mbrooke]

There are two formulas for voltage regulation, voltage regulation up and voltage regulation down. This is the formula for voltage regulation up.

I got this from another EE site:

Voltage regulation = IR cos phi + IX sin phi (pu) Cos phi=power factor of load

Technically this correct as PF does play a big part in it. Which I was planning on asking: Why does PF effect the regulation so much? What if the load was leading? Just wondering.



...................................................


Lastly: For those who work with the IEC- why are impedances typically higher? I'm not referring to them basing the impedance off the OFAF instead of the ONAN rating like the IEEE, but rather when adjusted for the ONAN rating, IEC designs are typically in the 10-12% region.

 

[mbrooke]

I'll start another thread if it works better- but what do they mean by Impedance volts? Why not just call it impedance?

 

[mbrooke]

For substations, we traditionally try to limit fault current to 10 kA and this is sometimes achieved by requesting a higher impedance design. You have to consider the lowside equipment ratings and this used to be 8-10 kA or less so you would not want to expose these devices to higher fault currents. If you have higher rated equipment you can lower the impedance accordingly, at least to the transformer design limit.

Well, you don't know who you're dealing with- cause I have a secret weapon for that


http://www.cooperindustries.com/con...s/resources/library/240_Fusing/CA132021EN.pdf

https://www.eaton.com/content/dam/e...-limiting-dropout-fuse-catalog-ca132027en.pdf



All kidding aside, yes, 10ka is typically what is sought for. Though there is always the temptation to use bigger units or operate the secondaries in parallel while breaking out the secret weapons :angel: