different currents on the parallel wires of a phase 2

[gaux]

hi, all i got a question that maybe is very easy for electric guys, we have a generator, 400 V 1500 A, and we use 6 wires to get out the current, and i got diferents currents on each wire, i tight them again but it´s not sucesful, and now i don´t know why...we cheek the terminals and all seems to be ok, does somebody know why...
the diference is about 10 % of the current....?

i hope somebody could help me cause i´m worried about some wire will burn.... thanks in advance

 

[Bronzeado]

Hi gaux,

Could you, please, post a picture showing the exact configuration of your 6 paralleled wires, including the distance among them?

Regards,

H. Bronzeado

 

[rbulsara]

That is pretty common, as the lengths of the parallel conductors are not exactly the same and slightest difference in impedance makes then share unequally.

 

[gaux]

yeah, i have no pic, but the distance is the same, they go to the same bus bar and in the other connection point, they goes to the same bar... what is the difference on distances that make a difference of 60 A.? the distance between generator and bus bars are 100 m.

 

[LionelHutz]

You'll never get the currents to be exact. 10% sounds decent for current sharing. I think the electrical code tells you to apply a derating factor for parallel cables to account for this.

I assume you are using a clamp-on meter. Try this for fun. Measure the current in each cable and sum these currents. At the same time, measure the total current. I bet they don't add up. You might see that just by putting the clamp-on around each cable you can change the impedance enough to affect the current.

 

[gaux]

But my question is why this difference? cause the lenght is more or less the same, maybe mutual inductance? could somebody explain this mechanical guy this?

 

[gaux]

and what does the electrical code says about this? cause i want to ask the guy that do the installation...cause now the supervisor didn´t accept us the instalation.....

 

[edison123]

Because even if the distances are identical (which is a big if), the resistances are not because of inherent material property. While the theoretical resistivity of copper is 1.7 micro-ohm.cm, it varies in the real world because of material impurities.

 

[Bronzeado]

Depending on the geometric configuration of the wires, the magnetic field ("impedance") for each wire may be different, so the current.

Herivelto

 

[tonyjony]

Mutual inductance could be the right answer, but Unbalanaced Loads is a far more common situation.

 

[tonyjony]

Sorry, I forgot the "wires of A PHASE"

 

[gaux]

maybe if the installer used a cheap wire? this theorical resistivity would change? do you think this is cause of resistivity of cooper?
and yes... i see the wires going out of the generators connection box and then go through the floor till the bus bars.... and seems to be the same, except that a difference of maybe 300 mm is enough for this difference.
in anyway i´m very interested in the electrical code, what they said about this, cause maybe ít have to be taken into account.

 

[gaux]

and about this.... about geometric configuration, what could be the expected difference in the worst situation? i don´t know if this could be mesasured or calculated?

 

[gaux]

ok. i forget to mention, i got 2 units else, and they are more or less shared, little diferences... 5 amps or so, but i don´t know if the manufacturer of the wire is the same... the section, yes.

 

[electricpete]

I think your question has been answered.

1 - As you increase wire size, inductive and mutual-inductive effects tend to become more important than resistive effects. Even if the lengths are identical, the wires on the outside of a group see different mutual inductance than wires on inside of a group (example of geometric configuration as discussed above).

2 - Clamp-on current measurements are not particularly accurate.

10% difference is not surprising.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Your question about difference in load sharing can be answered a number of ways. It is not particularly hard to model by finite element analysis if you provide a description of the configuration... how big are conductors, distance between the 6 conductors, arranged in a line or in a cluster, where is the ground plane, is there a shield?

With free program FEMM, the question can probably answered in 10 minutes. (Allow 8 hours to learn the program first!)

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I should clarify I am offering to run the program to simulate your configuration if you tell me the configuration.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[waross]

Forget resistance differences. If there is 1 meter difference in the length of conductors in a 100 meter run that is an average of 1/2% deviation or 1% difference in resistance. There is seldom that much difference in length and in this instance the deviation is much more than this.
There are two causes of this common issue.
1> Metering errors. The ammeter may be influenced by the magnetic fields of nearby conductors. When this is an issue, you may see different currents indicated at different locations on the same conductor. You may see a change in indicated current if you change the angle between the ammeter and adjacent conductors.
The other effect is impedance. The reactance of the conductors depends on their physical relationship to each other and to the ground plane. You may be able to reduce the unbalance by switching conductors. That is switch an "A" phase conductor with a "B" phase conductor. I am sorry but I can not suggest how to determine how to choose the conductors to be switched.
With six conductors in conduit there will be some connection combinations that will result in good current sharing and some combinations that result in poor current sharing.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[gaux]

120 mm per wire, and they go together by a cluster,

 

[electricpete]

120mm sounds awfully big. Maybe was supposed to be 12 mm ?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.