Proportional Increase of EGC If CM's Decrease?

[kenjsil]

This relates to increasing the EGC in proportion to the CM increase in conductor size under 250-122(b). (In CA, we're still using the '99 NEC, where this was only required for increases due to voltage drop, but I understand the 2002 code requires an EGC increase if the circuit conductors are increased for derating as well.)

If you have to increase the conductor size for derating or voltage drop and have to go from 1 conductor per leg to parallel conductors, you can have an increase in ampacity with a decrease in total CM of the conductors.

For example: 208Y/120 3 phase 4 wire feeder for 500A continuous nonlinear load. Standard OCPD is 700A. Suppose you're using 90C Cu wire with 75C terminals. Then the minimum conductor size for the load (at 75?C terminal rating) is 1500MCM good for 625A. But with a current-carrying neutral, you have to derate the 90C ampacity of 705A for 4 conductors at 80% = 564A. So you increase the conductors to 2 parallel 500MCM in 2 pipes. The total derated ampacity is 688A and using the next larger size OCPD, it works.

So, you've increased the total ampacity but decreased the total CM's from 1,500,000 to 1,000,000.

The EGC required by Table 250-122 for a 700A OCPD is 1/0 Cu.

Do you increase the EGC from this and by what calculation?

 

[don_resqcapt19]

Re: Proportional Increase of EGC If CM's Decrease?

Ken,
In my opinion, this installation does not require an increase in the EGC size. The EGC is sized so it can handle the available fault current. An increase in the circuit ampacity through the use of parallel conductors does not increase the available fault current. In fact it actually reduces the fault current due to the increase in impedance caused by the reduced amount of copper in the circuit conductors.
Don

 

[bphgravity]

Re: Proportional Increase of EGC If CM's Decrease?

I may be wrong in my thinking, but this is how I see the issue. The section specifically requires the increse in size proportional according to the circular mil area and not overall capacity.

So for your example, you have increased the capacity of the system but not the cm area of the system.

But what is the intent of the section. It is my understanding that the issue is overall circuit impedance and not capacity. Based on a basic calculation using the values of R in Table, the overall resistance of 1500kcmil is still less than the overall R of paralleled 500kcmil conductors. So its my feeling that since overall impedance of the system is not being increased, the EGC shuld not have to be increased even thought the system has a higher capacity.

 

[steve66]

Re: Proportional Increase of EGC If CM's Decrease?

I think you can forget about the single vs. parallel conductors. That seems to make it more complicated than it really is (but if I'm missing something here, I'm sure Don or Bryan will let me know Here is how I would calculate this;

1. 500A continuous requires 500 *1.25 = 625 amp wire

2. (2) sets of 400KCM with a 700A CB would be the smallest parallel conductors to satisfy this. Thus, upsizing the EGC would not be required for this. Per table 250.122, each pipe would have a 1/0 ground in it.

3. Due to derating, you need (2) sets of 500KCM.
Anything larger than (2) sets of 400 would require increasing the EGC size.

4. Finally, to calculate the EGC size, 500KCM/400KCM x 105600 CM = 132000 CM. A 2/0 would be required since it is 133KCM. (Use table 8 to convert 1/0 and 2/0 to CM or KCM.)

Steve

 

[steve66]

Re: Proportional Increase of EGC If CM's Decrease?

An increase in the circuit ampacity through the use of parallel conductors does not increase the available fault current.

Don:

I have to disagree with this statement. Less copper does not necessarly mean higher impedence. Due to the tendancy of current to flow on the surface of a conductor, more surface area usually equates to less impedance, a smaller voltage drop, and more available fault current. And parallel conductors usually provide more surface area than larger single conductors.

Steve

 

[charlie b]

Re: Proportional Increase of EGC If CM's Decrease?

I agree with Steve?s mathematic approach, and with his conclusion. In the original statement of the problem, Kenneth, you altered the problem mid-stream. You compared the ampacity of a single conductor with a set of two parallel conductors, and concluded that you have decreased the overall CM. That is mixing apples with oranges.

You should then have taken the 625 amp required capacity (i.e., from 500 amp continuous load), and multiplied it by 1.25 to obtain the new required ampacity. The result would be 781 amps. Since there is no single conductor with that ampacity, you would have to go back to the beginning. That would mean starting with a set of parallel conductors that are good for 625 amps, and then multiplying by the 1.25 factor. That is Steve?s approach.

 

[rbalex]

Re: Proportional Increase of EGC If CM's Decrease?

Kenneth,

Not that it seems to matter much, since it appears 90% of the jurisdictions in California don't know it, but the California Building Standards Commission (BSC) adopted the 2002 NEC in May of last year. See first paragraph of the second page of the following announcement on their website:

International Code Council (ICC) or National Fire Protection Association (NFPA) Building/Fire Codes (September 8, 2003)

The BSC has done a poor job getting the word out. Its just one more of the "issues" we're having here in California at the moment.

Since the 2003 "Triennial" adoption cycle was postponed a year, there are rumblings that the next "official" edition of the California Building Codes will use the 2005 NEC as the basis for Chapter 3.

The AHJ I'm working with now is still using the 1998 California Electrical Code which is based in the '96 NEC and swears its the only "legal" version. And for those of you who will jump in and suggest it may be by local adoption - it isn't. The state preempted local codes without specific BSC approval over 10 years ago and the city Charter specifically references "the latest California Building Code at the time of securing [the] permit." It was the version the project was originally permitted under; but the municipal codes permit (not require) the use of any subsequently adopted California Building Code(s). The only section of Code this AHJ knows is 90.4 - and he misinterprets that. In the four months that I have been involved with project the AHJ has yet to cite any relevant code accurately.

 

[charlie b]

Re: Proportional Increase of EGC If CM's Decrease?

Originally posted by steve66: Don: I have to disagree with this statement.

Sorry, Steve, but this time I have to disagree with you. Here is one example. I did not do a formal study, so I can?t say if this example would be typical of all cases.
</font>

• <font size="2" face="Verdana, Helvetica, sans-serif">Start with a 700 MCM uncoated Copper. Ampacity is 460 amps. DC resistance is .0184 ohms per 1000 feet.</font>

<font size="2" face="Verdana, Helvetica, sans-serif"></font>

• <font size="2" face="Verdana, Helvetica, sans-serif">Now look at a set of two parallel 4/0 uncoated Copper. Combined ampacity is the same 460 amps. DC resistance is one half of .0608, or 0.0304 ohms per 1000 feet.</font>

<font size="2" face="Verdana, Helvetica, sans-serif">By using the parallel conductors, you have increased the resistance, and therefore lowered the fault current that would be available at the end of the run.

Would this always be the case? I tend to think so. Resistance is given by a resistivity factor multiplied by length and divided by area. The resistivity is a constant. The parallel conductors would each be smaller in diameter than the single conductor, giving a smaller area, and therefore a higher resistance.

 

[don_resqcapt19]

Re: Proportional Increase of EGC If CM's Decrease?

Steve,
A 200' run of 600 kcmil has a 3 volt drop at 400 amps. Two sets of 3/0 in parallel have an ampacity of 400 amps. The voltage drop on the parallel conductors is 5.3 volts. The I^2R losses are higher, the impedance is higher and the available fault current is less with the parallel conductors.
Don

 

[steve66]

Re: Proportional Increase of EGC If CM's Decrease?

I did not do a formal study, so I can?t say if this example would be typical of all cases.

I didn't do one either, but it does seem like the numbers can go either way. I have a "Square D Motor Data" card. It has voltage drop per ampere per 100' for 1 phase, 3 phase, 95% PF and 80% PF. One can quickly find numbers to support either case for voltage drop.


But, fault calculations generally involve more than DC resistance. I compared 2 sets of 4/0 vs. one 750 KCM using a spreadsheet program and a published table of "Z" values. The 4/0 gives the higher fault current even if its DC resistance is greater. 11,725 amps for 750KCM, and 12,607 amps for 2 sets of 4/0.

Steve

 

[don_resqcapt19]

Re: Proportional Increase of EGC If CM's Decrease?

Steve,
I stand corrected. I did some short circuit calcs and found that often the 2 smaller conductors in parallel have a greater fault current at the end of the run than the single larger conductor. I'm not sure that I understand why. The voltage drop is one of the things that limits the fault current at the load end of the conductors and the voltage drop is almost always greater where parallel conductors are used.
Don

 

[bob]

Re: Proportional Increase of EGC If CM's Decrease?

The origional problem had a continuous load of 500
amps x 1.25 = 625 amps and a 700 amp OC.
Since it is necessary to use the 0.80 derating factor, the required ampacity of the conductor is
781 amps or 781/2 = 391 amps per conductor or
2 500 kcm per phase with an ECG of 1/0 cu. IMO this is the required conductor for this load. Now if you need to increase the conductor to 600 kcm per phase for VD, then you need to increase the EGC.

 

[charlie b]

Re: Proportional Increase of EGC If CM's Decrease?

I agree with Steve, in that the numbers could go either way. It is also clear that AC impedance and power factor would both play a role. It would take more of a study than I have time to perform to give a clear answer.

I disagree with Bob, in that derating for more than three current-carrying conductors is one of the ?increased in size? conditions that would cause us to increase the EGC. Two others are voltage drop and high ambient temperature.

 

[bob]

Re: Proportional Increase of EGC If CM's Decrease?

Charlie B I think you misread my post.
You said

I disagree with Bob, in that derating for more than three current-carrying conductors is one of the ?increased in size? conditions that would cause us to increase the EGC.

My point was

Since it is necessary to use the 0.80 derating factor, the required ampacity of the conductor is 781 amps or 781/2 = 391 amps per conductor or 2 500 kcm per phase with an ECG of 1/0 cu. IMO this is the required conductor for this load.

. I think the 0.80 is used to determine the minimum circuit capacity. I do not think this is an increase size that dicates the EGC size.
However if the temperature was over the 30C I would include it along with the adjustment factors in 310.15(B)2a as the minimum cirtcuit capacity and not consider it an increase size in conductor.
That said, I suspose the only thing left is voltage drop that would dictate an increase in
the EGC.

[ July 27, 2004, 08:48 PM: Message edited by: bob ]

 

[steve66]

Re: Proportional Increase of EGC If CM's Decrease?

I follow your logic, Bob, and the language in the code doesn't seem real clear. According to the 2002 NEC,

Where ungrounded conductors are increased in size

But it doesn't say increased from what size. I think we must assume that it means they are larger than the "code required size". But the code does require the 0.8 derating, thus, 500KCM would be the required size. So I think I agree.

But if we follow this logic, increasing wire size for ambient would also be a code requirement, so increasing the EGC for this wouldn't be required either.

That would leave us with what you said, an increase for voltage drop (since it is the only non-required increase) would be about the only time one would have to up-size the EGC.

One other thing: The increase in EGC size is required per 250.122(B). EGC size for parallel conductors are covered in 250.122(F).
I'm no longer sure the increase in size per (B) applies to parallel conductors. In effect, parallel EGC's are already upsized by virtue of the fact that there is a full size EGC in each conduit.

Steve

 

[charlie b]

Re: Proportional Increase of EGC If CM's Decrease?

Bob: I didn?t misread your post, I just disagreed with it. I still do. I note that Steve has changed his mind. I haven?t. Here is my viewpoint:

The required ampacity of the conductors is 625 amps. The question is, what size of conductor has that minimum ampacity. Or more to the point, what size of conductor will give you that ampacity with two conductors in parallel (i.e., 313 amps each)?

You could start by saying that a 400 MCM (THHN Copper) would answer the need, with its ampacity of 335 (at 75C). But then we note the requirement for de-rating. The ampacity of a 400 MCM (THHN Copper), de-rated for more than three current-carrying conductors, is found by taking 80% of 380 (the 90C value), and is 304 amps. That is not enough, since you need 313.

Next, the ampacity of a 500 MCM (THHN Copper), de-rated for more than three current-carrying conductors, is found by taking 80% of 430 (the 90C value), and is 344 amps. That is good enough.

Key Point: You had to increase the conductor size because the ampacity of the conductor got lower. Specifically, 310.15(B)(2)(a) states that this situation requires reduction in the ampacity of the conductors. It is not a matter of the required ampacity getting higher.

I reiterate my agreement with Steve?s original method and his original conclusion: That a 2/0 is needed here. I concede that the NEC wording is vague, but that is how I read it.

 

[charlie b]

Re: Proportional Increase of EGC If CM's Decrease?

Originally posted by steve66: I'm no longer sure the increase in size per (B) applies to parallel conductors. In effect, parallel EGC's are already upsized by virtue of the fact that there is a full size EGC in each conduit.

I don?t think we can take credit for an increase in size of the EGC on the basis that there are more than one of them. I think the purpose of having an EGC in each raceway is to safeguard against the possibility that a fault occurs within one of the raceways (with the fault, perhaps, being caused by a backhoe). You can get very nearly the full fault current (since the phase conductors are connected at the load end), and you need the full size EGC to carry it back to the source.

 

[bob]

Re: Proportional Increase of EGC If CM's Decrease?

Charlie B
I think one could make a case for your interpretation. As you and Steve said the code is not clear. In my 1999 handbook the section on this subject is listed as
250.122b Adjustments for Voltage Drop.
not so in the new code.

 

[steve66]

Re: Proportional Increase of EGC If CM's Decrease?

Posted by Charlie B..I note that Steve has changed his mind

In a way. I will still be up-sizing EGC's for any increase in conductor sizes. That's the safest thing to do. But I do think the NEC could improve the wording and make the intent a little clearer.

Steve

 

[kenjsil]

Re: Proportional Increase of EGC If CM's Decrease?

Thank you all for the responses. It is a privilege to receive the benefit of (collectively )several centuries of experience in the trade.

I realize that in this context comparing single and parallel conductors is apples and oranges. That was the reason for my post. As steve66 said: "But it doesn't say increased from what size". My humble interpretation was "the minimum conductor(s) for the load before any derating". In the case I posted, that would be a single 1500MCM conductor with a 75C ampacity of 625. The minimum parallel conductors are 2 x 400MCM with a total 75C ampacity of 670. That is already an increase. So it seemed to me that doing the CM-proportional increase calculation from 800000 was side-stepping the issue.

My answer is to resort to a fault-current and withstand-rating calculation in cases like this. I was hoping for something simpler. In my (admittedly limited) experience, an electrician presenting engineering references from outide the code can expect the inspector's 3rd degree.