paralleling feeders at different size?

Status
Not open for further replies.
A

anbm

Senior Member
Location
TX
Occupation
Designer
Can we run (2) different size paralleling feeders? i.e. 600kCM and 350kCM?
 
infinity said:
Why would you want to do this in the first place?
Click to expand...

I am thinking about to change a breaker from 600A to 800A and want to reuse existing wiring already pulled for 600A CB. Sound likes code does not allows this.
 
infinity said:
Were you looking to add a set of 350's?
Click to expand...

No, we already had (2) sets of 350 MCM for 600A CB, I want to add one more 3/0 set, the combined of these three sets will be okay with new 800A CB but not with NEC and try to figure why.
 
A plausible reason that the NEC does not permit conductors of different sizes to be placed in parallel because the current flow will not be safely divided between the different conductors. In the situation that you describe, using a 3/0 in parallel with two 350 kcmil conductors to carry 800A, the 350 kcmil conductors would carry more than their 'share' of the load.

When you have parallel conductors, the current divides up in proportion to the conductivity of the individual conductors. For conductors of the same length, the conductivity is proportional to the area. (Note: this is an approximation that ignores 'skin effect'; save that for another day.)

That 3/0 conductor has an area of about 168 kcmil. So if you put a 3/0 in parallel with a pair of 350 kcmil conductors, then the 3/0 will carry about 19.4% of the current, and the 350's will carry about 40.3% of the current. Take your '800A' figure and multiply it by those proportions. The 350's are left carrying more than 320A, the 3/0 is carrying less than 160A. The 800A breaker would not properly protect the parallel set of conductors.

-Jon
 
winnie said:
A plausible reason that the NEC does not permit conductors of different sizes to be placed in parallel because the current flow will not be safely divided between the different conductors. In the situation that you describe, using a 3/0 in parallel with two 350 kcmil conductors to carry 800A, the 350 kcmil conductors would carry more than their 'share' of the load.

When you have parallel conductors, the current divides up in proportion to the conductivity of the individual conductors. For conductors of the same length, the conductivity is proportional to the area. (Note: this is an approximation that ignores 'skin effect'; save that for another day.)

That 3/0 conductor has an area of about 168 kcmil. So if you put a 3/0 in parallel with a pair of 350 kcmil conductors, then the 3/0 will carry about 19.4% of the current, and the 350's will carry about 40.3% of the current. Take your '800A' figure and multiply it by those proportions. The 350's are left carrying more than 320A, the 3/0 is carrying less than 160A. The 800A breaker would not properly protect the parallel set of conductors.

-Jon
Click to expand...

A while back I had a job with parallel 500's for a service. We were required to measure the current on each conductor when we were finished. There was only a minor difference in the lengths due to the the radii of the bends in the conduit and the switchboard. I was amazed at the difference in the current loading.
 
anbm said:
No, we already had (2) sets of 350 MCM for 600A CB, I want to add one more 3/0 set, the combined of these three sets will be okay with new 800A CB but not with NEC and try to figure why.
Click to expand...

It would not be safe, the current would not divide the way you would need it to.

Cheapest way out is to add another set of 350s, the new set would have to be run the same way as the old set, same route, same length, same type of raceway.

Check out 310.4(1), 310.4(2), 310.4(3), 310.4(4) and 310.4(5)

If there is a equipment grounding conductor included that will also be a problem as the existing EGC will likely be sized for 600 amps not 800 amps and 250.122(F) requires a full size EGC in each pipe.
 
Lcdrwalker said:
A while back I had a job with parallel 500's for a service. We were required to measure the current on each conductor when we were finished. There was only a minor difference in the lengths due to the the radii of the bends in the conduit and the switchboard. I was amazed at the difference in the current loading.
Click to expand...
The length variations alone probably were not the only factors. When talking about sub-1-ohm resistances, each connection can also contribute to load-sharing imbalances; every bolt, lug, etc., should be checked again.

The closer one comes to maxing out paralleled-conductor ampacities, the greater the chance that even minimal imbalances can cause the one with the lowest impedance to be overloaded.
 
Status
Not open for further replies.
Top