running parallel wires

[LarryFine]

Would it really matter if the cables are the same length and the impedance of the cable was off by 10%?

That depends on how close to the edge the sizing was done, and how close to that number the actual loading becomes.

Length has no bearing as long as impedance is the same. But real world say's that it would be physically possible to have the exact same impedance unless you resort to significant figures.

One could easily exceed such a tolerance if connections are not made up equally tight and true, also.

P.S. Or, what Don said.

P.P.S Hey, the first post on a page showed up correctly! :grin:


Now, for one to attempt to adjust connections to compensate for conductor length variations would be nuts, in my opinion.

 

[varmit]

Forgive me for not being a physics wizard, but supposedly electricity flows at the speed of light (186, 282 miles per second ). At 60 hertz, this would be 3104.7 miles per cycle. At this speed, how could even a 50 percent difference in the length of paralleled 200 to 300 foot wire runs make a significant loading imbalance? I would think that any current differences between conductors of the same phase would be likely caused by termination variables or metallurgical differences.

No, I don't know everything and I am always interested in learning more.

Yes, I understand that the code requires the conductors of a phase to be the same length, and i try to do this, but is this based on science or: " This is how we have always done this."

 

[don_resqcapt19]

Varmit,
The speed is not the issue. The conductors are in parallel and create a current divider. The most current will flow on the wire with the least impedance. If there is too much difference in the impedance's of the conductors and the circuit is heavily loaded, you may have more current flowing in a conductor than is permitted.

 

[brian john]

Varmit,
The speed is not the issue. The conductors are in parallel and create a current divider. The most current will flow on the wire with the least impedance. If there is too much difference in the impedance's of the conductors and the circuit is heavily loaded, you may have more current flowing in a conductor than is permitted.

I still contend (in my experience) I see more current sharing issues with the routing of the conductors at each end then with issues with lengths of conductors. I THINK?

 

[LarryFine]

V, welcome to the forum!

Forgive me for not being a physics wizard, but supposedly electricity flows at the speed of light (186, 282 miles per second ). At 60 hertz, this would be 3104.7 miles per cycle. At this speed, how could even a 50 percent difference in the length of paralleled 200 to 300 foot wire runs make a significant loading imbalance? I would think that any current differences between conductors of the same phase would be likely caused by termination variables or metallurgical differences.

If we were talking about RF or video signals, such as RGB-HV, then matching lengths of multiple signal cabling is critical. But, not at 60 Hz.

At this low a frequency, the concern here is purely resistive, and how a small resistance difference can cause a big division of current-sharing imbalance.

For example, if you parallel two wires for, say, 500a, expecting each wire to carry 250a, and size it close, but there is, say, a 10% resistance difference.

Because of the division of current, one wire could see (in round numbers) 225a, and the other 275a. If you chose a 260-capable wire, it's overloaded now.

 

[varmit]

Varmit,
The speed is not the issue. The conductors are in parallel and create a current divider. The most current will flow on the wire with the least impedance. If there is too much difference in the impedance's of the conductors and the circuit is heavily loaded, you may have more current flowing in a conductor than is permitted.

Yes, I understand that the total impedance difference of each conductor, of each phase, is a factor in the conductor loading, however, in the wire sizes and lengths normally encountered in building electrical systems, it doesn't seem that the very small percentage wire length differences would have a significant impact.

On every paralleled circuit, that I have checked with an amp meter, there was a significant difference in conductor loading even if the wire length was as close as humanly practical and the terminations were identical. Could there be some little studied situation where the current flow may be more random than we would like to believe?

 

[varmit]

OK, my "50 percent" statement was a little off. I understand that the total impedance of the circuit affects the current, but I have checked several paralleled circuits with an amp meter where the wires were as close to equal length as humanly practical and were terminated in an identical fashion. On most of these there was a significant load difference (more than 10 %) between the conductors of phases. This seems almost a random choice of current path to the load. I had some time to waste once on a variably loaded parallel circuit. Using a non RMS meter- this was years ago, as the total load increased, the percent of imbalance decreased. it almost appeared that when the circuit was lightly loaded there may have been reflected harmonics that could have an effect on the meter reading.

 

[IslandSparky]

Yes, I understand that the total impedance difference of each conductor, of each phase, is a factor in the conductor loading, however, in the wire sizes and lengths normally encountered in building electrical systems, it doesn't seem that the very small percentage wire length differences would have a significant impact.

On every paralleled circuit, that I have checked with an amp meter, there was a significant difference in conductor loading even if the wire length was as close as humanly practical and the terminations were identical. Could there be some little studied situation where the current flow may be more random than we would like to believe?

That is why the NEC is so tight with load calculations. The assumption is always that the load factor is less than 80%.

 

[don_resqcapt19]

... Could there be some little studied situation where the current flow may be more random than we would like to believe?

Nope, but there are many things, other than the wire length, that have an effect on the impedance, including the routing of the conductors as stated by Brian in an earlier post.

 

[brian john]

That is why the NEC is so tight with load calculations. The assumption is always that the load factor is less than 80%.

It is?...............

 

[iwire]

however, in the wire sizes and lengths normally encountered in building electrical systems, it doesn't seem that the very small percentage wire length differences would have a significant impact.

So basically you are saying ohms law and other well established formulas have no meaning?

 

[infinity]

Do all of these ideas about exactly the same length have any real world concerns? Are parallel feeders burning up all over the place? The logic behind the requiment is valid but do we see any real problems out in the field when someone didn't measure the condcutors to the exact same length? With the exception of some short parallel secondary conductors I've never seen anyone worry about the length of parallel feeder conductors in the field.

 

[brian john]

Do all of these ideas about exactly the same length have any real world concerns? Are parallel feeders burning up all over the place? The logic behind the requiment is valid but do we see any real problems out in the field when someone didn't measure the condcutors to the exact same length? With the exception of some short parallel secondary conductors I've never seen anyone worry about the length of parallel feeder conductors in the field.

Occasionally we see some REAL WORLD issues, I'll post some IR pictures on Monday, as I mentioned before all of these were related to routing the conductors when terminating..

 

[LarryFine]

Are parallel feeders burning up all over the place?

Maybe they aren't because we've been at least trying to comply with the requirements.

 

[dbuckley]

Maybe they aren't because we've been at least trying to comply with the requirements.

I think that is very true.

Essentially, the current down paralleled cables divides as per the inverse of the resistance. The issue is at an edge condition, where you have two 700A capable cables paralleled for a 1400A load, and instead of a theoretical 700/700 split, you get a practical 500/900 split, and then at 80% loading you're at the edge of one cables rating.

So as long as everyone fails to grasp exactly whats going on, and just does what the book says, there we're all safe from parallel burndown...

 

[brian john]

Line side installed rather poorly

Load side 35' away, possibly a different electrician did these terminations? The fix was a reworking of the line terminations.

 

[kingpb]

Nice pics! Always worth a thousand words.

 

[rcwilson]

So basically you are saying ohms law and other well established formulas have no meaning?

Ohms law says current will divide in parallel paths based on the relative impedance of the two paths. Resitance is just a part of impedance. In a parallel circuit, the spacing of conductors in individual ABC bundles and orientation between bundles can make a big difference in current division, even if the resistances are exactly equal.

I have measured 25% loading difference in a 1200' run of 5-500kcmil per phase. We reterminated both ends and checked the terminations for high resistance joints with an IR scan under load. It was the orientation of the cables in tray that caused the imbalance. I had to lower the breaker setting to protect the cable until some loads were moved.

In the pictures above, you can see some cables are inside the bundle and some are on the outside. Their mutual indictance will be different leading to differnt impedances under load and different amp loading, even if they are exactly the same length.

 

[morepower]

Taken from Ferm?s Fast Finder Index?, play with this formula to see the effects of different lengths of parallels:

Lp*It/Lt = Ip

Where: Lp = Length of a parallel conductor

It = total current of the load

Lt = Total length (sum of parallels)

Ip = Current on a parallel conductor

I run across this problem an awful lot and it seems that few people understand the importance of the rule.

 

[don_resqcapt19]

Do all of these ideas about exactly the same length have any real world concerns? Are parallel feeders burning up all over the place? The logic behind the requiment is valid but do we see any real problems out in the field when someone didn't measure the condcutors to the exact same length? With the exception of some short parallel secondary conductors I've never seen anyone worry about the length of parallel feeder conductors in the field.

Probably the biggest reason we don't see the parallel conductors buring up is the very conserative load calculations in Article 220. If those calculations were more realistic, we probably would see more parallel conductor failures.