[Carultch]

NEC 250.122(B) requires that when I increase my current-carrying wire size to accomodate long distance voltage drops, that I must also upsize the ground wire proportionally in metal cross sectional area. I like to think in kcmils for this task.

For my examples, all wires are copper with 90C insulation and 75C terminals.

I understand it for single conductor per phase feeds. As a simple example, consider a 64A load on an 80A breaker that would normally have #4 wire, #8 ground, for local cases. Upsize the wire to #2 for voltage drop, and the proportional increase (160%) yields #6 Cu for ground.

What happens when your situation is on the cusp of either paralleling vs not paralling, and voltage drop causes the decision to parallel? Usually I consider 500 kcmil or 600 kcmil to be the maximum size before thinking about paralleling.

Example: a 400A breaker with a 320A load. Locally, it could have either 600 kcmil, or parallel 3/0, and in either case, a #3 ground per raceway. If a single raceway is shared, it would be parallel 4/0 because of the bundling derate.

Suppose I were upsize the live feeders to parallel 350 kcmil, for voltage drop. At this size of wire, if it is 3-phase AC, I wouldn't even about sharing a raceway. Dedicated raceway each set.

I understand that whatever the ground may be, each raceway must contain a full size. No dividing up the kcmil, as I originally would guess.

If 600 kcmil were considered the default, then I've upsized by a ratio of 7/6. Meaning the upsized ground would be #2.

If parallel 3/0 were considered the default, then I've upsized by a ratio of 2.08. Therefore the ground would be #2/0.

So which strategy is correct?

 

[HARRY305E]

NEC 250.122(B) requires that when I increase my current-carrying wire size to accomodate long distance voltage drops, that I must also upsize the ground wire proportionally in metal cross sectional area. I like to think in kcmils for this task.

For my examples, all wires are copper with 90C insulation and 75C terminals.

I understand it for single conductor per phase feeds. As a simple example, consider a 64A load on an 80A breaker that would normally have #4 wire, #8 ground, for local cases. Upsize the wire to #2 for voltage drop, and the proportional increase (160%) yields #6 Cu for ground.

What happens when your situation is on the cusp of either paralleling vs not paralling, and voltage drop causes the decision to parallel? Usually I consider 500 kcmil or 600 kcmil to be the maximum size before thinking about paralleling.

Example: a 400A breaker with a 320A load. Locally, it could have either 600 kcmil, or parallel 3/0, and in either case, a #3 ground per raceway. If a single raceway is shared, it would be parallel 4/0 because of the bundling derate.

Suppose I were upsize the live feeders to parallel 350 kcmil, for voltage drop. At this size of wire, if it is 3-phase AC, I wouldn't even about sharing a raceway. Dedicated raceway each set.

I understand that whatever the ground may be, each raceway must contain a full size. No dividing up the kcmil, as I originally would guess.

If 600 kcmil were considered the default, then I've upsized by a ratio of 7/6. Meaning the upsized ground would be #2.

If parallel 3/0 were considered the default, then I've upsized by a ratio of 2.08. Therefore the ground would be #2/0.

So which strategy is correct?

Any amendments in the 2014 mass code on this section?

 

[ponyboy]

I don't even see the question. I think you might be over complicating it.

 

[Carultch]

I don't even see the question. I think you might be over complicating it.

The question is "which strategy is correct?"

Single set 600 kcmil w/ #3 ground default, upsized to 2 sets of 350 kcmil with #2 ground each set.

OR

2 sets of 3/0 w/ #3 ground default, upsized to 2 sets of 350 kcmil with #2/0 ground each set.

Which strategy is correct for upsizing ground for voltage drop?

 

[Carultch]

Any amendments in the 2014 mass code on this section?

Good question, and applicable to about 80% of my projects, given that I am in Massachusetts. I haven't seen the 2014 amendments yet.

However, this project isn't in MA, so I'd like to know what the national rule on this is.

I don't know why, but somehow it is just my luck that I am always encountering borderline situations like this, when trying to interpret the code.

 

[HARRY305E]

I don't even see the question. I think you might be over complicating it.

He needs to increase the size of his ungrounded conductors because of voltage drop, In such a case he has to increase the size of his EGC according to 250.122 (B).

(B) Increased in Size. Where ungrounded conductors are increased in size from the minimum size that has sufficient ampacity for the intended installation, wire-type equipment grounding conductors, where installed, shall be increased in size proportionately according to the circular mil area of the ungrounded conductors.

 

[ponyboy]

I see what you're getting at but I'm not doing the math. What I'd do when I did this was find the percent of increase in the ungrounded conductors and apply it to the EGC. But to be perfectly honest, I just go right off of 250.122 and be done with it when it comes to parallel runs

 

[HARRY305E]

Good question, and applicable to about 80% of my projects, given that I am in Massachusetts. I haven't seen the 2014 amendments yet.

However, this project isn't in MA, so I'd like to know what the national rule on this is.

I don't know why, but somehow it is just my luck that I am always encountering borderline situations like this, when trying to interpret the code.

Here is a copy of the 2014 amendments.:thumbsup: http://www.mass.gov/eopss/docs/dfs/osfm/cmr/cmr-secured/527012.pdf

 

[HARRY305E]

Good question, and applicable to about 80% of my projects, given that I am in Massachusetts. I haven't seen the 2014 amendments yet.

However, this project isn't in MA, so I'd like to know what the national rule on this is.

I don't know why, but somehow it is just my luck that I am always encountering borderline situations like this, when trying to interpret the code.

The code does that to everybody,,,,:laughing:

 

[Carultch]

He needs to increase the size of his ungrounded conductors because of voltage drop, In such a case he has to increase the size of his EGC according to 250.122 (B).

(B) Increased in Size. Where ungrounded conductors are increased in size from the minimum size that has sufficient ampacity for the intended installation, wire-type equipment grounding conductors, where installed, shall be increased in size proportionately according to the circular mil area of the ungrounded conductors.

Exactly. And there are two different interpretations, depending on what "the minimum size that has sufficient ampacity for the intended installation" is considered to be.

If it said "the minimum size with fewest possible parallel conductors that has sufficient ampacity for the intended installation", then I'd know what to do. Obviously it doesn't say that, because it would be confusing to understand the simple single set case.

When you parallel because of local factors like amps, terminals, bundling and air temp, you usually get to use less kcmil of metal in your wire compared to the single set. It is easier to keep the temperature down on several little circles than on one big circle.

When you parallel because of voltage drop, it is usually the same total kcmil of metal either parallel or single set, once sized. The NEC resistance tables correspond (approximately) in a standard inverse proportional relationship between ohms/ft and kcmil, just like you would expect from the concept of Ohm's law and resistivity.

 

[HARRY305E]

Exactly. And there are two different interpretations, depending on what "the minimum size that has sufficient ampacity for the intended installation" is considered to be.

If it said "the minimum size with fewest possible parallel conductors that has sufficient ampacity for the intended installation", then I'd know what to do. Obviously it doesn't say that, because it would be confusing to understand the simple single set case.

When you parallel because of local factors like amps, terminals, bundling and air temp, you usually get to use less kcmil of metal in your wire compared to the single set. It is easier to keep the temperature down on several little circles than on one big circle.

When you parallel because of voltage drop, it is usually the same total kcmil of metal either parallel or single set, once sized. The NEC resistance tables correspond (approximately) in a standard inverse proportional relationship between ohms/ft and kcmil, just like you would expect from the concept of Ohm's law and resistivity.

There should be examples in chapter 9 on this very question.

 

[Carultch]

There should be examples in chapter 9 on this very question.

I'll check Chapter 9 when I get in to work tomorrow, and let you know.

 

[Carultch]

There should be examples in chapter 9 on this very question.

Well, I checked chapter 9, and they don't seem to cover it.

I'm gonna go with basing it on the minimum local size (with fewest possible parallel sets). Meaning #2 Cu ground, would be my answer.

This seems to make the most intuitive sense to me. A single #3 Cu is required to carry the fault current and trip the breaker on a short distance, and when length increases by a factor of 1.6 beyond the maximum length of the single wire option, the ground's resistance per unit length per raceway needs to go down by a factor of 1.6.