208/120v 3P4W Neutral CCC

[arc_flashed]

I have seen this brought up in a few other threads so I apologize for any overlap, but I am still confused and hoping someone can help me out with understanding.
I have a 75kva 208/120vac wye transformer inside a facility. The question comes up with quantity of current carrying conductors and the wire sizing from the secondary of this transformer to the distribution panel. Does the neutral count as a current carrying conductor between the transformer secondary and the 200amp main breaker on my panel? Standard 75degC would make this a 3/0 conductor unless you apply the 80% for 4 CCCs.
I understand a 3-phase load with a neutral would not count under 310.15(E)(1) (2024 edition of NFPA) but my panel is made up of 100% single phase loads (all receptacles).
Since all my loads are 1P (H, N, G = 2CCCs) and connected to the panel's neutral bus, wouldn't my transformer X0, and therefore my neutral conductor, carry all of this current and need to be counted as a current carrying conductor pushing me to an 80% conductor rating?

 

[LarryFine]

No, because the neutral is still only carrying the difference current. If your three lines happen to be loaded perfectly equally, there is zero neutral current in the feeder.

For any and every reduction in current in any line, which would increase neutral current, there is that much less line current in one or more ungrounded conductors.

The aggregate current will never exceed the equivalent full line current.

 

[winnie]

Since all loads are single phase 120V, all the current returns to the neutral bus.

At the neutral bus, current from different phases will add in a vector fashion, so that the total current returning on the neutral wire to the transformer will be lower than the phase currents.

If the loads on each phase are perfectly balanced, then the neutral current will be zero.

Unless something strange is going on (such as all 120V non-linear loads), the neutral in a 3 phase 4 wire 208/120 system does not count as a CCC.

Jon

 

[wwhitney]

At the neutral bus, current from different phases will add in a vector fashion, so that the total current returning on the neutral wire to the transformer will be lower than the phase currents.

This answer is correct in practice, and the OP should ignore this abstract pathological digression:

Seems like if the load A-N is purely inductive load (current 90 degrees lagging) and the load B-N is purely capacitive load (current 90 degrees leading), while the load C-N is purely resistive, then for one choice of A-B-C rotation, the A-N and B-N current vectors are each only 30 degrees off the C-N current vector. So if they are equal in magnitude, their vector sum is (1 + sqrt(3)) times their individual magnitude. Yes?

A similar thing would happen with single phase split with the A-N and B-N loads as in the previous example. The current vectors would be in phase and add, rather than subtract.

But for practical loads with power factors not so far from 1, then the vector addition should give you a current of magnitude not larger than any of the summands.

Cheers, Wayne

 

[winnie]

Yup, you can always create pathological loading schemes that break commonly expected balance.

-Jon

 

[Carultch]

Does the neutral count as a current carrying conductor between the transformer secondary and the 200amp main breaker on my panel? Standard 75degC would make this a 3/0 conductor unless you apply the 80% for 4 CCCs.

To summarize when the neutral counts as a CCC:
1. If it is a mandatory part of the return path, even for a condition of balanced loads. E.g. a phase-to-neutral circuit, where its easiest to identify this. Less obvious: 120/208V 3-wire circuit derived from a 2-pole breaker on a 120/208V wye system, because it requires the neutral to carry the full current so that (Ia + Ib + In) will add up as vectors to zero.
2. If the loads are harmonic intensive, particularly triplen harmonics, such that multiples of 180 Hz frequencies add up rather than cancel, on the neutral conductor.

Other than that, neutral doesn't count as a CCC when:
1. It is only used for instrumentation purposes, and carries negligible current.
2. It is only used for carrying the imbalance, given a possibility of unbalanced, that are close enough to linear. It turns out that no matter what the imbalance, it still doesn't generate more heat than a balanced case at full load. Ohmic heating is proportional to current^2. Since Ia^2 + Ib^2 + Ic^2 + In^2, can never be greater than Ia^2 + Ib^2 + Ic^2, no matter what the imbalance, the neutral need not count as a CCC.

 

[wwhitney]

2. It is only used for carrying the imbalance, given a possibility of unbalanced, that are close enough to linear. It turns out that no matter what the imbalance, it still doesn't generate more heat than a balanced case at full load. Ohmic heating is proportional to current^2. Since Ia^2 + Ib^2 + Ic^2 + In^2, can never be greater than Ia^2 + Ib^2 + Ic^2, no matter what the imbalance, the neutral need not count as a CCC.

Did you mean "power factor close enough to 1" rather than "close enough to linear"? Because if I understand correctly capacitors and inductors are linear, and the pathological example I posted contradicts the inequality in your last sentence.

Cheers, Wayne

 

[topgone]

No, because the neutral is still only carrying the difference current. If your three lines happen to be loaded perfectly equally, there is zero neutral current in the feeder.

For any and every reduction in current in any line, which would increase neutral current, there is that much less line current in one or more ungrounded conductors.

The aggregate current will never exceed the equivalent full line current.

BTW, harmonics throws curved balls to that usual idea. Triplen harmonics are additive at the neutral, hence it is possible to have the neutral current higher than that of the phase currents.

 

[LarryFine]

Since Ia^2 + Ib^2 + Ic^2 + In^2, can never be greater than Ia^2 + Ib^2 + Ic^2, no matter what the imbalance, the neutral need not count as a CCC.

That's what I said.

The aggregate current will never exceed the equivalent full line current.

 

[LarryFine]

BTW, harmonics throws curved balls to that usual idea. Triplen harmonics are additive at the neutral, hence it is possible to have the neutral current higher than that of the phase currents.

Very aware of that, but actually achieving that point is not easy.

I thought the simpler answer was more appropriate for the O.P.

 

[JoeStillman]

It's in The Code...

310.15(B)(5) Neutral Conductor.
(a) A neutral conductor that carries only the unbalanced current
from other conductors of the same circuit shall not be required
to be counted when applying the provisions of 310.15(B)(3)(a).
(b) In a 3-wire circuit consisting of two phase conductors
and the neutral conductor of a 4-wire, 3-phase, wye-connected
system, a common conductor carries approximately the same
current as the line-to-neutral load currents of the other conductors
and shall be counted when applying the provisions of
310.15(B)(3)(a).
(c) On a 4-wire, 3-phase wye circuit where the major portion
of the load consists of nonlinear loads, harmonic currents
are present in the neutral conductor; the neutral conductor shall
therefore be considered a current-carrying conductor.

 

[jim dungar]

BTW, harmonics throws curved balls to that usual idea. Triplen harmonics are additive at the neutral, hence it is possible to have the neutral current higher than that of the phase currents.

But this is very rarely a concern in real life. About the time it made it into the NEC the world was moving to switched mode power supplies for small single phase loads. These draw relatively small bursts of current so it is typically hard to have enough on general circuits and transformers without some mutual cancellation.

My experience has been triplen harmonics are hard to predict in advance. So 200% neutrals and non-linear transformers are specified simply as a CYA, especially when few transformers are actually fully loaded which is why the DOE has set 35% as the point where a transformer's efficiency is determined.

So be aware that this phenomenon exists.

 

[wwhitney]

Triplen harmonics are additive at the neutral, hence it is possible to have the neutral current higher than that of the phase currents.

A simple example of this phenomenon is the case that the current is a square wave instead of a sine wave.

Take a sine wave, shift it 120 degrees and also 240 degrees, add all three up, and you get zero. For a square wave, do the same, and you get a square wave of 3 times the frequency and the same magnitude. [You can do this addition graphically, or by noting that if you divide the period into 6 equal segments, each of the 3 square waves are constant on each of the 6 segments, so just can add them up piecewise.] So for 3 identical loads L-N that draw square wave current, the neutral current would be the same magnitude as that on any of the three legs.

The harmonic analysis gives you the same result: the Fourier decomposition of a square wave consists of each odd harmonic n with magnitude 1/n. Then if you consider just the harmonics that are odd multiples of 3, the sum of those components has a magnitude that is 1/3 of the square wave. Those triplen harmonics are the ones that add on the neutral instead of canceling, and with 3 such loads the neutral current is of the same magnitude as one of the square waves.

But as a practical matter, are there many loads that draw current with a harmonic content close to that of a square wave? Because that seems pretty extreme.

Cheers, Wayne

 

[winnie]

But this is very rarely a concern in real life. About the time it made it into the NEC the world was moving to switched mode power supplies for small single phase loads. These draw relatively small bursts of current so it is typically hard to have enough on general circuits and transformers without some mutual cancellation.

My recollection is slightly different.

I believe the issue reared its head when switching power supplies started becoming common because of computer equipment, and then the issue became less relevant as the switching power supplies were improved to reduce triplen harmonics on the input current.

Roughly, switching power supplies rectify the input power to DC, and then use transistor switching of that DC to develop the desired output voltages. The input current harmonic content is controlled by the design of the rectifier circuit. Simple diode rectifiers with capacitor ripple filters will draw current only at the peak of the AC voltage, creating a harmonic rich current flow. More advanced rectifier designs can greatly reduce the harmonic current drawn by a switching power supply.

Here is a case report from 1995 detailing a neutral overload issue caused by harmonics:

The case of the soon-to-be overloaded neutral conductor.

Odd triplen harmonic currents and neutral overloading go hand-in-hand.A facility manager of a data center was informed by the Information Systems Department that additional file...

www.ecmweb.com

-Jon

 

[jim dungar]

My recollection is slightly different.

I believe the issue reared its head when switching power supplies started becoming common because of computer equipment, and then the issue became less relevant as the switching power supplies were improved to reduce triplen harmonics on the input current.

Roughly, switching power supplies rectify the input power to DC, and then use transistor switching of that DC to develop the desired output voltages. The input current harmonic content is controlled by the design of the rectifier circuit. Simple diode rectifiers with capacitor ripple filters will draw current only at the peak of the AC voltage, creating a harmonic rich current flow. More advanced rectifier designs can greatly reduce the harmonic current drawn by a switching power supply.

Here is a case report from 1995 detailing a neutral overload issue caused by harmonics:

The case of the soon-to-be overloaded neutral conductor.

Odd triplen harmonic currents and neutral overloading go hand-in-hand.A facility manager of a data center was informed by the Information Systems Department that additional file...

www.ecmweb.com

-Jon

That, almost 3 decade old, article is about an expected problem, based on limited information.
If that had been my facility, I would have measured the harmonic content of the existing loads and then tried to estimate the impact of future loading. SCR line notching produced much more harmonics than other types of rectifiers.
I am not saying triplen harmonics don't cause problems, just that they are not as prevalent as manufacturers of 'solutions' say in their literature, kind of like the way instructors always warn new electricians about skin effect even though it is negligible for 60Hz systems especially with conductors smaller than 250kCMIL.

 

[don_resqcapt19]

That, almost 3 decade old, article is about an expected problem, based on limited information.
If that had been my facility, I would have measured the harmonic content of the existing loads and then tried to estimate the impact of future loading. SCR line notching produced much more harmonics than other types of rectifiers.
I am not saying triplen harmonics don't cause problems, just that they are not as prevalent as manufacturers of 'solutions' say in their literature, kind of like the way instructors always warn new electricians about skin effect even though it is negligible for 60Hz systems especially with conductors smaller than 250kCMIL.

Most everything I read about that "problem" was written by someone who had an economic interest in helping me solve that "problem".

 

[winnie]

That, almost 3 decade old, article is about an expected problem, based on limited information.

Agreed. They did have the actual measurements of current, including on the neutral.

I stated (but should have emphasized) that the article was from nearly 30 years ago, specifically because the problem appeared (in very limited cases) and then was solved by better front end rectifiers and standards requirements to reduce harmonic components in switching power supplies.

Jon

 

[topgone]

That, almost 3 decade old, article is about an expected problem, based on limited information.
If that had been my facility, I would have measured the harmonic content of the existing loads and then tried to estimate the impact of future loading. SCR line notching produced much more harmonics than other types of rectifiers.
I am not saying triplen harmonics don't cause problems, just that they are not as prevalent as manufacturers of 'solutions' say in their literature, kind of like the way instructors always warn new electricians about skin effect even though it is negligible for 60Hz systems especially with conductors smaller than 250kCMIL.

Basically, that's what we did in trying to have a complete picture of our systems. We commissioned a vendor to do the harmonics frequency scan of our substations and the data enabled us to calculate the state of things of our substation transformers, because we have so many VFD loads and rectifiers. It helped us a lot. We are now sure of how much an expansion is possible with our existing setups.