Splitting large ampacity need across multiple smaller conductors - pros/cons (?)

[nateholt]

If anyone has time, please take a look at this and let me know if this is a common industry practice...

Large motor 500 feet away, 3ph motor (1.25 x FLA) = 535A, continuous duty, no ambient temp de-rating required for wiring.

From tables, 1000MCM yields 545A ampacity for 75C insulation. Assuming the three phase conductors in one conduit, no conductor de-rating required so this should work.

But maybe cheaper to go with 3x250MCM rated at 255A non-de-rated ampacity?

That would give 9 current carrying conductors in the conduit? meaning de-rate ampacity by 0.7 to 178.5A each. Times 3 gives just enough to cover the full motor?s (1.25 x FLA) value of 535A.

The additional advantage of the smaller conductors is that the voltage drop would be less. For the 1000MCM at 0.85 PF and 535A for 500 feet, I ran the full calculation using Table 9 reactance, AC resistance, and sine of arccos (messy), came up with 18.3 volts drop. But running the same calculation on the 250 MCM wires gives voltage drop of only 11.3 volts.

So, splitting the load over three smaller conductors saves 25% of the metal and cuts the IR losses in the wiring by 35%. The negative might be that the conduit could be bigger and not sure the cost of 3 -250MCM is that much less than the cost of 1-1000MCM (?).

Is the above logic consistent with industry practices? Any insight appreciated.

Thanks,
Nate Holt.

 

[charlie b]

I have a VD calculator that does not include 1000 MCM conductors. But at 750 MCM (not enough ampacity, I know), I get a VD result of 8 volts. Certainly, with 1000 MCM, the VD would be lower. So I do not confirm your calculated values of voltage drop. With a set of three 250 MCM conductors, I get about the same result of 8 volts.

As to derating, you get to use the 90C column as the base value for ampacity. Three sets of 250 MCM conductors would derate from 290, not from 255. So the total ampacity would be 3 times 290 times 70%, or 609 amps.

Is there an option of running separate conduits? You could do this job with two sets of 300 MCM conductors.

 

[coulter]

I cam tell the electricians will really like you if you spec 350 or 250 instead of 1000.

Working 1000kcmil is about like bending a short, fat bolt.

carl

 

[rcwilson]

Is Parallel Less Expensive?

Is Parallel Less Expensive?

When I run the numbers from the estimating guides, 1000 kcmil wire is usually the highest installed cost in $ per amp, even including the cost of running added parallel conduit runs. As a rule, I parallel anything over 500 kcmil.

I never trust the estimating guides and rely on the electrician's who actually do the work. The feedback I get is the ease of pulling 2 - 350's saves costs over pulling 1000's.

Another way of looking at it is weight= labor. Heavier wire & larger conduit means more weight = more labor = more costs. The smaller, parallel wire runs usually have less pounds per foot of circuit.

I'm interested in hearing what the people who work with it every day prefer.

 

[winnie]

The IR drop in the 1000 kcmil conductor will be lower than that of 3x 250 kcmil conductors.

However once you get to the bottom of table 9, the reactive impedance of the conductors is of similar magnitude to the resistance.

It is plausible to me that the IZ of 3 x 250 kcmil would be lower than the IZ of a single 1000 kcmil. However I would not trust the table 9 values for Z to apply to your situation. The actual reactive impedance will depend upon the geometry of the conductors in the conduit, and with 9 conductors the geometry will certainly be different from what table 9 assumes.

-Jon

 

[nateholt]

Thanks for the quick reply.

This isn't for an actual job, my question is just theoretical. My background is EE but I've been in the software industry for a long time now, and am pretty ignorant on this real-world stuff. Why can I go with the 90C column even if the wiring is rated at 75C?

I've played with various VD calculators and, again speaking from a newbie point of view, many seem flawed when dealing with large diameter conductors. It seems the "skin effect" and power factor are not taken into account... and these two factors seem to really take hold on VD calculations when the conductor size grows large. Most seem to just go with a "k=12.9" factor for copper. I tried the calculation listed in NEC Table 9 Note #2 for large diameter conductors and that is what yielded the higher VD value in the posting.

One on-line calculator that seems to to avoid this k-factor limitation and go with VD calculation based upon actual reactance / AC resistance / PF factors is the "Southwire" one at

http://www.southwire.com/voltagedropcalculator.jsp

Anyway, I need to sort this out, and your response is much appreciated!

thanks again
Nate.

 

[BryanMD]

Another way of looking at it is weight= labor. Heavier wire & larger conduit means more weight = more labor = more costs. The smaller, parallel wire runs usually have less pounds per foot of circuit.

I'm interested in hearing what the people who work with it every day prefer.

Take a look at the 20 amper's thread.
Same basic issue just scaled down.

http://forums.mikeholt.com/showthread.php?t=96803

 

[charlie b]

Why can I go with the 90C column even if the wiring is rated at 75C?

You can?t. Perhaps an error on my part to assume you were talking about using conductors with 90C insulation systems. But that is what I deal with most of the time.

If, and only if, the insulation system is rated for 90C, then you can use the 90C column of Table 310.16 for the purposes of derating. This applies both to derating for ambient temperatures over 30C and derating for more than three current carrying conductors in a raceway. The place in the NEC that says you can do this is the last sentence of 110.14(C).

 

[nateholt]

Anyone / Everyone...

A few more questions about splitting up a large ampacity need into multiple conductors.

1. What is "rule of thumb" wire-size breakpoint where you would consider going with multiple conductors?

2. What is the maximum number of conductors that you would split across? For example, let's say you have ampacity and voltage drop that calculates out to 1-1000MCM, 2-350MCM, 3-250MCM or 5-4/0 for each phase. Would you ever split into 5 conductors? What is the "normal" max you would normally go?

 

[spsnyder]

I'm not sure "rule of thumb" is a good term. I would say that the size for large ampacities would be 250 to 500 MCM with as many runs as required for the total ampacity. I also do not try to spec greater than 500 MCM based on feedback I got from the guys in the field. I use separate conduit also if I can. Other feedback was that setting up multiple spools adds to the difficulty.

 

[nateholt]

Thanks for the reply. And if you run multiple conduits to accommodate the split-parallel conductors, you would make sure that each conduit carries conductor representatives of all three phases (i.e. you wouldn't run one phase per conduit), right?

 

[480sparky]

Thanks for the reply. And if you run multiple conduits to accommodate the split-parallel conductors, you would make sure that each conduit carries conductor representatives of all three phases (i.e. you wouldn't run one phase per conduit), right?

Parallel conductors are covered in and 300.3(B) and 310.4.

 

[dbuckley]

535A at 500ft, I'd suggest you try an transformer calculation, might be cheaper to go 4160V (or higher?) for the run. Especially if the facility is nearby fed with MV.