Skin effect in Generator power cables at 120Hz 4

[Jensdsds]

Hello,

I have some questions about skin effect.
Information to know:

Large Generator size MW.
Power cables at 300mm^2 single core.

When a large single core cable like the 300mm^2 cable goes from 50Hz operations to 120Hz, should i dimension it with new current reduction factors?
I have searched the internet, standard, forums etc. for information on this issue.
Usually an increase in frequency from 50Hz to 120Hz shouldn't be a problem, considering skin effect.
But in large cables the skin depth becomes a problem, going from 9.22mm at 50Hz to 5.95mm at 120Hz.
The 300mm^2 cable have a diameter of 24mm, and a 400mm^2 cable have a D of 29mm.

From my simulations an increase in frequency from 50Hz to 120Hz in a 300mm^2 cable will create a 20% increase in cable loss.
Should this increased loss not be considered when dimensioning cables.

Thanks in advance!

 

[jpts]

Hi
Single core cables are still made from bunch of smaller diameter strands (so it is not like one thick solid conductor). If the diameter of the strands is smaller than skin depth (5.95mm) skin effect should not be a problem. I assume this is the case. Often the strands are less than 1mm in diameter

Furthermore, skin effect in free air is quite negligible compared to coil conductors which are inside stator iron core.

My company makes MW electric motors/generators, and never have taken this into account in power cable selection, even though frequencies are often close to 300Hz. But with coil conductors inside slot it is different story

Hope this helps

 

[Jensdsds]

Hi,

Thank you for your answer!
It is good to know that you've never experienced any problems up to 300Hz. Are your cables sizes also in the range of 185-400mm^2?

Am i wrong to believe that the strands inside the conductor should have individual insulation, to reduce skin effect completely? Like the litz wire
From my teachings the stranding of the conductor is mainly to increase functionality, bending radius, handling etc.

During your work with MW motor/generators, have you tried measuring temperature difference in the cables, at different frequencies?
Thanks in advance!

 

[jpts]

We have experience up to 240mm^2 cables. Yes, stranding is done for functional reasons (to give flexibility like you said), not to minimize skin effect (strands are not insulated either)

We mostly use cables with very small strand diameter (such as Siwo-Kul from Nexans) since they are very flexible. Diameter of single strands is only couple of tenths of mm.

We have not measured temperatures at different frequencies. But I think the main point is that if the skin depth is much bigger than strand diameter, you will not have any issues with skin effect. I guess this should be the case now even at 120Hz?

We have also experience on Litz wire machines (now that you mentioned it). Our experience is that so-called bare Litz-wire (without any strand insulation), or oxidized Litz wire works really well too (meaning that there is no enamel on strand conductors).

 

[waross]

From my simulations an increase in frequency from 50Hz to 120Hz in a 300mm^2 cable will create a 20% increase in cable loss.

Cable ampacity is determined to a large extent by internal heating versus heat rejection.
Good engineering practice will be to derate the cable to recognize the increased losses.
Skin effect is a magnetic effect and magnetism does not respect insulation.
I have always suspected that litz wire was effective due to the separation of the conductors due to the thickness of the insulation and the inverse square effect of magnetic flux with increased distance.

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

 

[7anoter4]

The permissible current rating calculated according to IEC Standard 60287-1-1 depends on a few factors as thermal resistance of the insulation ,jacket and other as duct or conduits , ground or air and the conductor resistance.
I=SQRT((Tc-Ta)/R/K) Tc conductor maximum temperature Ta ambient temperature
K=a constant function of insulation and jacket thickness and other thermal resistances.
Whether the same cable will remain installed under the same conditions and only the frequency will increase the allowable current will depend only on the increase of the resistance, i.e., inversely proportional to the square root of the resistance ratio.
For instance, for 300 mm^2 copper-if we neglect the proximity effect-the derating factor will be 1/1.035 -only for skin effect.
Usually, for cross section area above 1600 mm^2 copper only, the conductor is segmental. See for instance:

https://onlinelibrary.wiley.com/doi/pdf/10.1002/eej.20576

 

[7anoter4]

If the single core cables run on trefoil formation, the 400 mm^2 cable -low voltage 75oC XLPE insulated- the derating factor-including skin and proximity effect-for 120 Hz with respect to 50 Hz is 0.927 and for 300 mm^2 0.952.

 

[jpts]

Hi Bill,
You can think Litz wire working with same principle as Roebel bars. The main idea is that the single atrands are transposed/twisted along the conductor. Then the voltages induced by slot leakage will be equal in all strands, cancelling out circulating currents due to skin effect. Without this transposition, there would be voltage difference between top and bottom part of the coil conductor, causing circulating currents. This would be comparable situation to normal round-wire windings.

 

[Jensdsds]

Thanks you all for the discussion, and especially 7anoter4 for the interesting paper!

I hope you have a great day.

 

[7anoter4]

For copper conductor up to 500 mm^2- since the error is less than 0.5% -you may use IEC 60287-1-1 formula 2.1.2
For larger cross section area, you have to use formula no.1 from the above article.
In order to calculate ber(x),bei(x),ber'(x) and bei'(x) see:

https://acwhyte.droppages.com/kelgenpaper.htm

 

[Jensdsds]

Thank you for your answers 7anoter4.

I am not sure if this is an option, but would it be possible for you to share the formula 2.1.2, as i currently doesn't have access to the standard.

Thank you in advance

 

[7anoter4]

Attached is Article 2.1.2

 

[7anoter4]

Also attached Table 2

 

[Jensdsds]

Thanks a lot!