Interesting 2 Section Panel connection

[iwire]

The raceways themselves have the same characteristics. The PB is not a raceway. IMO, the requirement regarding the raceways stops at the point they enter the panelboards. I still don't especially like it.

Raceway '1' starts in cabinet 'A' and ends in cabinet 'B'.

Raceway '2' starts in cabinet 'A' and ends in cabinet 'C'.

IMO Those raceways do not have the same characteristics as required by 310.4 no matter how much loose interpreting we want to do.

 

[hardworkingstiff]

Method 1 has a problem in that the parallel conductors for each section are not the same length. Each section has to add the length of the "jumper" conductors for the other set of wires thereby causing a violation of 310.4(1).

Don, you are correct in your statement that no installation can meet this requirement, but I believe this to be a flagrant violation.

If each section had a 200-amp overcurrent device for that individual section, electrically I would see no problem with this installation. Since each section is a MLO, I can see one section overloading the set of conductors that landed directly on its bus.

Did the contractor have any manufacturers instructions?

 

[hardworkingstiff]

I am curious. What is this sectional panel used for? Are both sides rated 400 amp each?

I've never dealt with a sectional panel, but I would suspect that 400-amps would be a small load to sectionalize. If each section is rated 400-amps, then I would be concerned the parallel 3/0's are not enough.

 

[infinity]

I am curious. What is this sectional panel used for? Are both sides rated 400 amp each?

I've never dealt with a sectional panel, but I would suspect that 400-amps would be a small load to sectionalize. If each section is rated 400-amps, then I would be concerned the parallel 3/0's are not enough.

Yes, both sections would be rated for 400 amps. This is actually a very common installation in commercial work.

Why wouldn't 400 amps of capacity be enough? If the calculated load required a 400 amp feeder why should we assume that it would be inadequate?

 

[don_resqcapt19]

Bob,

Raceway '1' starts in cabinet 'A' and ends in cabinet 'B'.

Raceway '2' starts in cabinet 'A' and ends in cabinet 'C'.

IMO Those raceways do not have the same characteristics as required by 310.4 no matter how much loose interpreting we want to do.

What if they are in the same enclosure, but with a divider in the enclosure between the conduit entries? I don't see how running the conduit to the second tub changes the physical(electrical) characteristics of the run.
Don

 

[don_resqcapt19]

petersonra,

Parallel conductors shall be the same length. For purposes of this section, parallel conductors shall be deemed to be the same length if the lengths of the parallel conductors do not vary in excess of 1% between the shortest and longest conductor, but in no case to exceed 5 feet.

I don't see any need for the 5' part of that proposed rule. The current divison is inversely related to the length and the 1% length change will only result in a 1% current difference between the conductors. If the run in 5,000' long the 5' part would be almost impossible to comply with. I think that the code does need a rule like this as the current rule would let an overly agressive inspector red tag the installation for 1".
Don

 

[hardworkingstiff]

Why wouldn't 400 amps of capacity be enough? If the calculated load required a 400 amp feeder why should we assume that it would be inadequate?

Of course you are correct, if the calculated load is 400-amps or less, they would be sufficient.

I still think method 1 is wrong for the reasons stated.

 

[infinity]

I think that the code does need a rule like this as the current rule would let an overly agressive inspector red tag the installation for 1".

How could the current rule possibly be enforced unless it was extremely obvious that the conduit runs are different. For example one run goes the opposite way around a floor from the other(s). I don't see how an inspector or the installer for that matter, could prove or disprove that the conductor lengthes are equal.

 

[hardworkingstiff]

Most of my parallel installations are from the POCO transformer (UG service) to my panel. Usually an 800-amp or 1000-amp service. I take a lot of care to route the conduits so they are as close to the same length as possible. I buy all the wire cut the same length, and then check it on site prior to pulling it in.

I do everything I can to make sure they are all the same length. I try to be there when the POCO makes up the transformer so I can ask them to make sure they cut the same amount off of each set of wires. They (POCO employee) nod their head yes and act as if they understand, then do whatever they damn well please.

 

[Polarcat]

Have been doing this a long time and would never install par. feeders in that manner, do not know what it vialatets (gray area at it's finest), but it just dosent sit well with me. I would install only as per example #2.

 

[LarryFine]

The raceways themselves have the same characteristics. The PB is not a raceway. IMO, the requirement regarding the raceways stops at the point they enter the panelboards. I still don't especially like it.

Raceway '1' starts in cabinet 'A' and ends in cabinet 'B'.

Raceway '2' starts in cabinet 'A' and ends in cabinet 'C'.

IMO Those raceways do not have the same characteristics as required by 310.4 no matter how much loose interpreting we want to do.

Technically speaking, each run does have the same characteristics:

Panel B's feeders look the same from that panel's perspective as Panel C's feeders do from its perspective. Each feeder feeds one panel directly and the other through the jumpers. They are identical to one another. That's not a defense, however; I feel that it's still wrong.

Nobody has really brought up performance issues. If the phases in the two panels are unevenly loadced, and differently, the jumpers could easily carry currents that might not balance out, and can create eddy currents that cause the nipple to heat up. That's a definite no-no.

 

[iwire]

Technically speaking, each run does have the same characteristics:

Yes they do...... if you leave out the fact that they do not go to the same place.

The destination is as much a characteristic of the raceway as the number of bends, the lengths, the material.

Could you fully describe the raceway run without describing the start and end point? :?:

 

[hardworkingstiff]

What if they are in the same enclosure, but with a divider in the enclosure between the conduit entries? I don't see how running the conduit to the second tub changes the physical(electrical) characteristics of the run.
Don

If they were in the same enclosure and the manufacturer had a bus that went through (or around) the divider, then I would agree method 1 would be ok.

Since the contractor is installing the conductors between the two buses, then I think method 1 is wrong unless the manufacturer specifically prints this is an allowed methodology (I'd really be surprised to see that in print).

 

[Pierre C Belarge]

In method one, what happens if one of the jumpers was to fail?
I am not comfortable with that installation.

Method two is not a problem that I can see.

 

[petersonra]

In method one, what happens if one of the jumpers was to fail?

What happens if any single conductor fails in a parallel conductor system?

 

[LarryFine]

Could you fully describe the raceway run without describing the start and end point? :?:

(First of all, we're all in agreement that this is just a friendly, fun, rainy-Sunday dialogue, right?)

Yes, I could, but obviously don't need to. The right fix here would be to use the original Method 2, with one difference: If I'm not mistaken, with both runs landing in one panel, the jumpers would need to be the same capacity as the feeders, since either panel can individually be loaded beyond 200 amps.

In my opinion, the other way to fix this is both to replace the 400a OCP with two 200a OCP's, and remove the jumpers.

 

[petersonra]

Method 1 has a problem in that the parallel conductors for each section are not the same length. Each section has to add the length of the "jumper" conductors for the other set of wires thereby causing a violation of 310.4(1).

The jumpers are clearly not part of the parallel conductors. They are part of the termination.

 

[petersonra]

Nobody has really brought up performance issues. If the phases in the two panels are unevenly loadced, and differently, the jumpers could easily carry currents that might not balance out, and can create eddy currents that cause the nipple to heat up. That's a definite no-no.

The problem with this argument is that you it assumes that with a load coming from the LH side PB, all the current comes from the LH side conductors (LH and RH referencing the drawing supplied).

That is just plain not the case.

 

[hardworkingstiff]

The jumpers are clearly not part of the parallel conductors. They are part of the termination.

Each side is rated 400-amps. The jumpers are rated 200-amps. Either side fully loaded will be using the jumpers as conductors (of course this is just my opinion). I sure would like to see the manufacturers recommended installations practice.

 

[LarryFine]

I suppose I was thinking on an infinitesimal scale.

Any difference between the feeders' impedances would cause current flow in the jumpers, and if said imbalance was different between the panels' respective phases, that current would flow imbalanced through the nipple.

Am I out of trouble yet? :roll: