Using 200% Neutrals

[DetroitEE]

Recently at a meeting among engineers at my firm, the topic of 200% neutrals came up. Common practice seems to be that you provide a 200% neutral for panelboards that serve a majority of non-linear loads, such as computer power supplies. However, one person made a comment that computer power supply technology has advanced over the years, and that they do not produce as many harmonics as they once did.

The main point of our discussion was to determine if we still feel it is necessary to provide 200% neutrals for large computer loads. Does any one know if there's any truth to computer power supply harmonics being less of an issue now? When do you guys typically use a 200% neutral?

 

[LarryFine]

I never have. While harmonics may cause triplen currents, the normal shared-neutral current still keeps the overall neutral current to a minimum.

 

[cripple]

Using 200% Neutrals

The NEC does not address the 200% sizing of neutral it does addresses the neutral to be considered as a current carrying conductor if load consists of nonlinear loads, harmonic currents are present.
But you might want to check the type of transformer being use, if it is K-rated transformers check with the manufacture.


http://www.lanecoburn.com/images/Neutrals Pure Power Fall 2005 K Rated Transformers R.pdf

 

[wirenut1980]

I have never seen the need for a 200% neutral in my experience. I am sure some cases probably exist, but they are rare.

 

[bob]

I never have. While harmonics may cause triplen currents, the normal shared-neutral current still keeps the overall neutral current to a minimum.

Larry
I think You are saying that the normal non-harmonic currents cancel and in some way affects the harmonic neutral currents. I don't believe that is correct but I may not be understanding your point. Please expand your thoughts.

 

[Dennis Alwon]

I never have. While harmonics may cause triplen currents, the normal shared-neutral current still keeps the overall neutral current to a minimum.

What about systems that don't use mwbc as may be the case with computer centers, etc. I agree 200% is crazy but I have heard of oversized neutrals.

 

[Besoeker]

I never have. While harmonics may cause triplen currents, the normal shared-neutral current still keeps the overall neutral current to a minimum.

I have come across situations where a 100% rated neutral was overheating as a result of non-linear single phase loads.
I have commented on it previously.

 

[zbang]

I saw this once maybe 16 years ago, the load was almost entirely 2kva UPSs and some directly connected PCs. By shuffling the UPSs, we got a good balance and the neutral current dropped to a sane value. With newer equipment, it should be even less of a problem.

 

[LarryFine]

Larry
I think You are saying that the normal non-harmonic currents cancel and in some way affects the harmonic neutral currents. I don't believe that is correct but I may not be understanding your point. Please expand your thoughts.

Only that the neutral must carry only normal imbalanced currents, so only the harmonics are of real concern. It's not like the neutral must carry full 'normal' current in addition to harmonics.

If only one or two phases had large non-linear loads, they would contribute little to heating the neutral. The harmonic current is the worst when the loads are most completely balanced.

 

[LarryFine]

I have come across situations where a 100% rated neutral was overheating as a result of non-linear single phase loads.
I have commented on it previously.

I saw this once maybe 16 years ago, the load was almost entirely 2kva UPSs and some directly connected PCs. By shuffling the UPSs, we got a good balance and the neutral current dropped to a sane value. With newer equipment, it should be even less of a problem.

It certainly and obviously can and does happen. I'm just saying that I haven't experienced it. But, then, I do tend to be conservative when it comes to loads of these types (commercial computers, power supplies, etc.) when I'm designing.

 

[ron]

With the advent of pofer factor corrected power supplies on IT equipment about 5-8 years ago, I haven't done 200% neutrals since then.

 

[mivey]

I would be very suprised to find a need for a 200% neutral. In fact, I would go so far as to say you will never see a need for one. I would buy up to 150% but to the best of my recollection, about 173% is the most you might ever see in extreme situations and that would be rare.

Just my recollection from some past reading on the topic.

 

[suemarkp]

Also realize that you can use a neutral conductor to 100% of its rating continuously (assuming it is not terminated on a circuit breaker), whereas you're limited to 80% for the ungrounded conductors of the same circuit. So that buys you some margin right there. At most, I'd go up one size in a neutral.

 

[DetroitEE]

Thanks for all the input guys.

Maybe I should have used the term oversized neutral as opposed to 200% neutral. I think 200% is just a typical way to size an oversized neutral (i.e. two grounded conductors of the same size in lieu of one), so maybe that's throwing people off.

I think Ron's response is along the lines of what I was looking for; I'll have to look into power factor corrected power supplies a bit more.

 

[LarryFine]

I think 200% is just a typical way to size an oversized neutral (i.e. two grounded conductors of the same size in lieu of one), so maybe that's throwing people off.

Still can't parallel smaller conductors.

 

[roger]

What about systems that don't use mwbc as may be the case with computer centers, etc. I agree 200% is crazy but I have heard of oversized neutrals.

Dennis, that would take care of the branch circuit neutrals, but the feeder neutral could still see problems HOWEVER, I am in the camp that says the additive harmonic paranoia is highly overblown.

Roger

 

[dbuckley]

Back where I used to live, and back in the day, 200% neutrals within data centre distribution used to be the norm, but as Ron noted above, computer power supplies are less damaging to the supplies than they used to be, so massive oversizing is generally no longer required.

With theatrical lighting, oversized neutrals still a good idea (and called for in the theatrical articles). With camlock cables getting a 200% neutral was easy, you just used two neutrals, along with one ground and one hot per phase, so you had six wire camsets. But the difficulty of providing oversized neutrals in standard multicore cales has caused the rule makers in the UK to think differently, and now you are required to downrate your distribution's capacity to match what your neutral can do.

 

[Dennis Alwon]

Dennis, that would take care of the branch circuit neutrals, but the feeder neutral could still see problems HOWEVER, I am in the camp that says the additive harmonic paranoia is highly overblown.

Roger

I thought that was my point. The feeders would still see the harmonics but 200% is unlikely.

 

[jrvazzer]

200% Neutral

200% Neutral

In he 90's all our computer loaded electrical panels had 200% neutrals and later our fluorescent ltg panels with electronic ballasts. The computer branch circuit were required to have #10 neutrals. We work on Jp Morgans, where there desk computers which was hundreds per floor were circuited to a 75KVA Harmonic transformer and the panels on the secondary side was 200% neutrals. Prior to this our transformers were very hot and harmonics was the culprit. Maybe now is not and issue but back then neutrals were over loaded, transformers were very hot including lighting circuits neutrals.

 

[iwire]

If you design the circuits and feeders conservatively from the start I see no need for larger neutrals.