Why isn't large wiring used more often?

Status
Not open for further replies.
J

Junior_EE

Member
Location
New York City
Typically, for commercial buildings, we do not specify any copper feeders larger than 600 kcmil or aluminum feeders larger than 750 kcmil. Once we reach these limits, we begin running multiple sets of conductors to distribute load.

I am an electrical engineering manager for an MEP consulting firm. One of my employees asked, for a project we're working on, why we couldn't increase a 600 kcmil copper feeder to a 700 kcmil feeder to account for excessive voltage drop. For one, I know that it's our company standard to move to 2 sets of conductors beyond 600 kcmil copper. Why do most buildings not have copper wire larger than 600 kcmil?

In the past, fellow engineers have given me incoherent ramblings about conductors' weight, bending radius, difficult to work with, not a standard size, etc. Can someone please help me better understand these reasons? I would prefer to hear from electricians, but would be delighted to hear engineers' understanding from past projects.
 
Junior_EE said:
Typically, for commercial buildings, we do not specify any copper feeders larger than 600 kcmil or aluminum feeders larger than 750 kcmil. Once we reach these limits, we begin running multiple sets of conductors to distribute load.

I am an electrical engineering manager for an MEP consulting firm. One of my employees asked, for a project we're working on, why we couldn't increase a 600 kcmil copper feeder to a 700 kcmil feeder to account for excessive voltage drop. For one, I know that it's our company standard to move to 2 sets of conductors beyond 600 kcmil copper. Why do most buildings not have copper wire larger than 600 kcmil?

In the past, fellow engineers have given me incoherent ramblings about conductors' weight, bending radius, difficult to work with, not a standard size, etc. Can someone please help me better understand these reasons? I would prefer to hear from electricians, but would be delighted to hear engineers' understanding from past projects.
Click to expand...
1. Large conductors become very inefficient in ampacity per circular mill. Compare 3 times the ampacity of 250 to a single 750

2. Getting terminals for larger than 500 can be a bit of a hassle sometimes. Sure factory ordered gear will come with pretty much whatever you want (as long as you spec'd it, default is typically 500's) but can be a problem with off the shelf stuff.

3. Harder to work with.
 
You get more ampacity with less material due to the skin effect of A/C. Take for example 500 CU which has an ampacity of 380 @ 75 degree. Now look at 2 250 CU which have an ampacity of 510. You get 25% more ampacity with the same amount of CU in this example.
 
Last edited:
Another way to look at this is to compute the cost per amp based on CMill size. A nice size for parallel is 3/0 as its 200 amps
 
We go as high as 750 kcmil copper and that's the limit. Anything larger than that is nearly impossible to work with and the labor cost is higher than the savings if you used smaller conductors.
 
texie said:
You get more ampacity with less material due to the skin effect of A/C. Take for example 500 CU which has an ampacity of 380 @ 75 degree. Now look at 2 250 CU which have an ampacity of 510. You get 25% more ampacity with the same amount of CU in this example.
Click to expand...
Skin effect is part of it, but the ratio of surface area to conductor area is, I think, as significant. Heat must escape the conductor to the environment.
 
texie said:
You get more ampacity with less material due to the skin effect of A/C. Take for example 500 CU which has an ampacity of 380 @ 75 degree. Now look at 2 250 CU which have an ampacity of 510. You get 25% more ampacity with the same amount of CU in this example.
Click to expand...

When ampacity governs, it is the surface area to volume ratio, that makes more parallel sets of smaller conductors, get more amps/kcmil. This is why a 400A circuit built with 600 kcmil, could also be built with 424 kcmil by using parallel 4/0's. That is about a 30% reduction in the total amount of copper. Ampacity depends on the wire's ability to discharge heat out through the electrical insulation and conduit, that is generated within the conductor. The disadvantage is that the voltage drop will be greater, so this is generally an advantage for short length circuits.

When voltage drop governs, on a first order analysis, it is generally is the total kcmil that matters, regardless of the number of parallel sets. In DC, the skin effect doesn't even happen. In AC, the skin effect is a second-order factor that comes in to play, and does give you an advantage to paralleling what you otherwise could do with a single larger conductor. You will be slightly more efficient at curtailing voltage drop with 2 sets of 300 kcmil, than the alternative of an individual 600 kcmil. This advantage is nowhere near as significant as the advantage of paralleling, when ampacity governs. In this example, the benefit is 1.6% less voltage drop. I.e. it would make a 2% voltage drop become 1.968%, by switching from 1x600 to 2x300's.
 
Last edited:
Carultch said:
When ampacity governs, it is the surface area to volume ratio, that makes more parallel sets of smaller conductors, get more amps/kcmil. This is why a 400A circuit built with 600 kcmil, could also be built with 424 kcmil by using parallel 4/0's. That is about a 30% reduction in the total amount of copper. Ampacity depends on the wire's ability to discharge heat out through the electrical insulation and conduit, that is generated within the conductor. The disadvantage is that the voltage drop will be greater, so this is generally an advantage for short length circuits.
Click to expand...
And with very short parallel conductors, a little more voltage drop on the conductors can be beneficial for keeping the currents in the conductors more evenly balanced. That's because the conductor resistance is much better controlled than connector contact resistance,etc. which may not be completely negligible when voltage drops are very small.
In some applications such as when paralleling power transistors they even insert small "ballast" resistances to improve current sharing.
 
Up to 350 MCM we figure almost the same labor hours. Above that you can almost figure on going up on every size increase for 500, 600, 750, etc. Plus fitting size goes up dramatically plus cost/Amp for the cable is going up to the point where multiple conductors per phase is cheaper even with the extra raceway cost. That’s assuming you can find things. Above 500 the standard lugs in breakers for instance aren’t that large so it requires bus bars or power distribution blocks or lugs to make splices to switch from say 750 to 350 or 500. So adding tremendous work to an otherwise easy job.

Usually the electricians just say that anything over 350-500 just shows that a whoever designed it didn’t know what they were doing. Just because it’s on a table doesn’t mean it’s a good idea. Kind of like using 400 MCM. Yes it exists but it’s like using 5” plumbing...just don’t do it.
 
Status
Not open for further replies.
Top