Feeders length for parallel runs must be within 3% length for both runs of one phase...?

[EJPHI]

I was told to change how I ran my wire for parralel runs of feeders. What's the code requirement for running parralel runs for feeders does the length need to match within 3%?

Sent from my Moto Z (2) using Tapatalk

A lot of good comments.
From another perspective consider that copper resistance has a temperature coefficient of + 0.4%/degree C. Two parallel wires with a length difference of 3% will share current in proportion to their resistance. The shorter one will carry more current, and with enough current, will warm up by about 8C before the resistance and therefore the current is equal in both wires. This 3% length match guideline seems reasonable.

Having all the short wires in one conduit and all the long ones in the other is not much of a concern from a resistance perspective because additional heating from all the short wires bundled together in the same conduit will cause their resistance to go up and their portion of the shared current to go down. Sort of a negative feedback effect.

The inductive effect from the different parallel wire lengths is probably not much concern. The longer wire has more inductance but the shorter wire, with sufficient current will warm up to balance the current to compensate for both resistance and inductance.

EJPHI

 

[jusme123]

You've got better eyes that I do. I've never seen a cabinet screwed together... just welded.

......most commercial/industrial panels come with top and bottom panel plates screwed on. The top plate is pre punched with ko’s and the bottom is blank, if your running conduit, you can switch the panels top plate and bottom plate and make custom ko’s to match pipe rack into panel.

 

[paulengr]

The motor's imbalance tolerance is a separate discussion, actually.

Multiple conductors of any given phase will be forced to have the same voltage at both ends by the terminals. Current imbalance is the concern, and the closer to ampacity they happened to have been sized, or rather loaded, the more that matters.



You two seem to be contradicting each other.

My point was that differences in length between PHASES in three phase loads matter due to voltage drop. Obviously single phase doesn’t. Your point was that with multiple conductors per PHASE it matters but it’s a current (impedance) issue and I’d agree, having seen it first hand along with not following proper phasing among multiple conduits (inductive heating).

Not sure why if I saw “few inches” in ten feet differs from “3.6 inches” or few feet at 100 feet contradicts 30 feet at 1000. Same issue. And in either case it does not matter until you approach an ampacity limit at which point you hit the “limit” faster on one cable. So if one cable is 1.0 ohms and the other is 1.03 ohms the parallel combination is 0.508 ohms. Say the current is 100 A then the voltage drop is 50.8 V. The current will be 51 A on one cable and 49 A on the other. That’s not much but it can be enough to push one cable over the ampacity limit before the others. This sounds impossible but trust me, industrial plants are really good at running their equipment at 99% of full load at times.

 

[kwired]

......most commercial/industrial panels come with top and bottom panel plates screwed on. The top plate is pre punched with ko’s and the bottom is blank, if your running conduit, you can switch the panels top plate and bottom plate and make custom ko’s to match pipe rack into panel.

For Square D NEMA1 enclosures like OP has in picture there is no top/bottom but often will have KO's in one end and not the other. Pick which one you want to be top and install panel components accordingly.

 

[five.five-six]

Code just says they have to be the same length with no specified tolerance. Arguably the 3% spec is looser than what the NEC allows.

The point of parallel feeders is that current divides between the various paths so that you can add the ampacity of the paths. Uneven length means uneven current flow, thus one of the parallel paths could overheat.

-Jon

In theory, the lower impedance Conductor will heat up under load making it the higher impedance conductor.

Then again, in theory, there is no difference between practice and theory. In practice, there is.

 

[winnie]

In theory, the lower impedance Conductor will heat up under load making it the higher impedance conductor.

Then again, in theory, there is no difference between practice and theory. In practice, there is.

This is a good point to keep in mind. For copper and aluminum conductors, the resistance increases with temperature, tending to compensate for small differences in conductor length. I think that the shorter wire will run hotter, but will still be the lower impedance conductor, but the difference will be less than the length difference.

Contrast this to other situations where things with negative temperature coefficients are put in parallel. In particular LED emitters should not be in parallel without ballast resistors, because they become more conductive as they heat up. Quickly one LED can hog the current and let its smoke out.
Jon