Definition of “Voltage to Ground”

[charlie b]

I wanted to gather opinions on this issue, before posting anything in the “Proposals for the Next Cycle” forum.

NECA and ECM are continuing the very good work started by Charlie Trout (may he rest in peace) in publishing a daily code question and response. I started to compose an email to contest today’s response. But I looked at the NEC itself first. It turns out that the response correctly described the NEC language, although I had not expected that to be the language in the code. Charlie’s Rule saved me from embarrassing myself. Here is the issue:

The article 100 definition of “Voltage to Ground” provides a separate meaning for grounded and ungrounded systems. For the ungrounded systems, the definition of “Voltage to Ground” is “the greatest voltage between the given conductor and any other conductor of the circuit.” I think this is wrong. I think that the voltage between any point in an ungrounded system to planet Earth can be any value whatsoever. It’s not limited to the system’s line-to-line value. I would prefer that the phrase “Voltage to Ground” be left undefined, with no more than a brief mention that that voltage value is not constrained to be any specific value.

Am I misinterpreting the intent of this definition?

 

[jim dungar]

You may be on the right track.

From an engineering stand point we know the voltage to ground is based on thing like leakage current, coupling capacitance and ferroresonance. This voltage can easily exceed the L-L values, particularly during arcing events. it is also possible that in a balanced system the voltage is no higher than what might be seen on a solidly grounded system.

 

[don_resqcapt19]

Since is it undefined and can be anything, setting the voltage to ground on an ungrounded system to the phase to phase voltage makes sense to me. There are rules in the code based on the voltage to ground, and if you make the voltage on this system undefined, you have to rewrite all of the rules that used the term voltage to ground.

 

[infinity]

The definition of voltage to ground correlates with Table 110.26(A)(1).

 

[jim dungar]

Since is it undefined and can be anything, setting the voltage to ground on an ungrounded system to the phase to phase voltage makes sense to me. There are rules in the code based on the voltage to ground, and if you make the voltage on this system undefined, you have to rewrite all of the rules that used the term voltage to ground.

In my career I have fielded many questions from electricians that measured 277V L-G on a 480V ungrounded system and were confused because it wasn't 480V.

 

[LarryFine]

From an engineering stand point we know the voltage to ground is based on thing like leakage current, coupling capacitance and ferroresonance. This voltage can easily exceed the L-L values, particularly during arcing events.

Let's not forget the very real danger of a primary-to-secondary fault.

 

[jim dungar]

There are rules in the code based on the voltage to ground, and if you make the voltage on this system undefined, you have to rewrite all of the rules that used the term voltage to ground.

I could live with a definition that said something like: 'for the purposes of this code, the L-G voltage shall be considered as equal to the maximum L-L voltage'.
This would be similar to what was done with the 'neutral' definition for 240/120V 3phase 4 wire systems.

 

[ggunn]

I could live with a definition that said something like: 'for the purposes of this code, the L-G voltage shall be considered as equal to the maximum L-L voltage'.
This would be similar to what was done with the 'neutral' definition for 240/120V 3phase 4 wire systems.

But in that case the line to neutral voltage is line to ground. How would you distinguish actual line to ground voltages in grounded neutral systems from line to line voltages? I am not in favor of calling something what it's not.

 

[jim dungar]

ggunn,

When you have an ungrounded 480V circuit what voltage rating do you specify for the standoff insulators for the bussing in a switchboard? How did you arrive at that value?

In a 240/120 3phase 4 wire system the common point on one winding is not an equal voltage to all of the line conductors. However, the NEC went out of its way to define it as the 'neutral point' anyway. As Don said, there is a lot of stuff in the NEC that needs a defined L-G voltage.

 

[LarryFine]

I believe the use of L-L voltage as the maximum for L-G voltage for an ungrounded system is based on the assumption that maximum L-G voltage (not based on induced voltages or accidental HV contact) will occur during a L-G fault on one phase.

 

[ggunn]

ggunn,

When you have an ungrounded 480V circuit what voltage rating do you specify for the standoff insulators for the bussing in a switchboard? How did you arrive at that value?

I dunno; I've never had to do that. I'm just saying that calling L to L voltage L to G is just not correct when the second L is not grounded, and if there is a grounded N present it just introduces ambiguity. Call it what it is.

 

[jim dungar]

I believe the use of L-L voltage as the maximum for L-G voltage for an ungrounded system is based on the assumption that maximum L-G voltage (not based on induced voltages or accidental HV contact) will occur during a L-G fault on one phase.

Absolutely.
However, many electricians expect to see 480V on an ungrounded system, probably due to their training, so when they measure 277V L-G they assume the system is solidly grounded.

 

[don_resqcapt19]

Absolutely.
However, many electricians expect to see 480V on an ungrounded system, probably due to their training, so when they measure 277V L-G they assume the system is solidly grounded.

Sounds like a training issue more than a code issue. "90.1(A) ...This Code is not intended as a design specification or an instruction manual for untrained persons."

 

[jaggedben]

...

Am I misinterpreting the intent of this definition?

What Don said. The intent of Article 100 definitions is not to describe the physical world precisely. The intent is to provide a basis for applying rules in the code. For example 250.97 uses 'voltage to ground' as the basis for requiring certain types of bonding, or not. The rule could not be sensibly applied to ungrounded systems if the code were to define voltage to ground for ungrounded systems as 'any value whatsoever.' Before proposing a change to the definition it would behoove one to research every instance instance in the code where 'voltage to ground' is used to apply a rule, and to solidly research whether changing the definition would have any cost-effective benefit to safety.

 

[jaggedben]

I dunno; I've never had to do that. I'm just saying that calling L to L voltage L to G is just not correct when the second L is not grounded, and if there is a grounded N present it just introduces ambiguity. Call it what it is.

Whether you remember noticing or not, you've certainly dealt with ungrounded DC circuits in the 400+ voltage range, where the voltage to ground as defined by the code caused certain code requirements to kick in.

 

[ggunn]

Whether you remember noticing or not, you've certainly dealt with ungrounded DC circuits in the 400+ voltage range, where the voltage to ground as defined by the code caused certain code requirements to kick in.

Of course, but + to - isn't + to G or - to G unless it's the second ground fault on the opposite polarity conductor. One does, also of course, have to take that into account.

 

[jim dungar]

Sounds like a training issue more than a code issue. "90.1(A) ...This Code is not intended as a design specification or an instruction manual for untrained persons."

You are correct, it is likely a training issue.
But, aren't most electricians primarily trained by other electricians? Don't many, if not most, electricians turn to the NEC when they want something defined?

 

[wwhitney]

s/voltage to ground/effective voltage to ground/g

Problem solved. : - )

Cheers, Wayne