Parallel conductors

[don_resqcapt19]

Don’t you have to have the same amount and same size of conductors in each raceway for parallel runs though? You can’t put 4 conductors in 1 pipe and 8 in the other can you?

If they are connected in parallel, that would be correct. There is no reason to connect all of the conductors together at the load end. You can easily run three sets of conductors in the two raceways to feed the three loads. The conductors would only be connected together at the transformer so they are not parallel sets of conductors.

 

[infinity]

If they are connected in parallel, that would be correct. There is no reason to connect all of the conductors together at the load end. You can easily run three sets of conductors in the two raceways to feed the three loads. The conductors would only be connected together at the transformer so they are not parallel sets of conductors.

I agree. Since the poster mentioned that this if from a one line drawing I would assume that the designer has something in mind when he called for parallel conductors. Maybe a tap for future expansion. Could even be a typo.

 

[kwired]

I agree. Since the poster mentioned that this if from a one line drawing I would assume that the designer has something in mind when he called for parallel conductors. Maybe a tap for future expansion. Could even be a typo.

I'd say the designer made a mistake somewhere. OP indicates parallel conductors, which from a transformer to a tap can ahead of some meter sockets would typically be done that way, but then specifies three sets of 350 but there is only two raceways to run this in.

We don't know the actual load so we can't really speculate if two sets of 500 is sufficient, though it probably is. This is a condition where even though we have the multiple 400, 200 and 100 amp disconnects, these conductors can be sized to the load and if actual load is only say 500 amps then only two sets of 350 would be fine unless voltage drop were a concern.

 

[infinity]

I'd say the designer made a mistake somewhere. OP indicates parallel conductors, which from a transformer to a tap can ahead of some meter sockets would typically be done that way, but then specifies three sets of 350 but there is only two raceways to run this in.

The mistake is the bold part. Everything after that is just speculation, like if the 3 sets of 350's are copper that would provide more ampacity than the 400+200+100 amp meters would ever need.

 

[Julius Right]

If the cables are parallel connected -from one busbar to another-in 2 conduits of 4", it will be better to use 4*4*350 mcm-2*4*350 in each conduit and this will full fill all NEC recommendations.

 

[kwired]

If the cables are parallel connected -from one busbar to another-in 2 conduits of 4", it will be better to use 4*4*350 mcm-2*4*350 in each conduit and this will full fill all NEC recommendations.

Does that give all parallel components equal impedance so the current will divide equally? Doesn't seem like you can rely on it being equal if not all conditions of each parallel conductor are not as identical as possible.

 

[Julius Right]

See 310.10 (H) Conductors in Parallel.(3) Separate Cables or Raceways.

 

[don_resqcapt19]

See 310.10 (H) Conductors in Parallel.(3) Separate Cables or Raceways.

If you are trying to say that permits 4 conductors in one raceway and 8 in the other, that is not correct.

The language about "one paralleled set" apply to the set of conductors used for a single phase. That means the A phase sets do not have to match the B phase sets.

 

[kwired]

If you are trying to say that permits 4 conductors in one raceway and 8 in the other, that is not correct.

The language about "one paralleled set" apply to the set of conductors used for a single phase. That means the A phase sets do not have to match the B phase sets.

I'm in agreement here. Placing 3 phase A conductors in one raceway and 2 in the other of a parallel setup can lead to overall impedance differences and cause more current to flow on some the individual conductors than others which can lead to overloading of those individual conductors even though overall load is still under the intended ampacity of the parallel set. You have to keep as many characteristics of each individual conductor of a set as similar as possible. Phase A and Phase B are different sets and don't have to be identical but all of conductors of A need to be same, all of B needs to be same and so on.

 

[Julius Right]

I said 8 conductors in each conduit-2*4*350 in each conduit.

 

[kwired]

I said 8 conductors in each conduit-2*4*350 in each conduit.

Ok for some reason reading it again looks different than it did before. Initially thought you had a different combination in one raceway than the other and apparently Don saw it the way I did as well.

 

[winnie]

If you have 2 4” parallel conduits going from a transformer to a tap-can to feed a 100amp, 200amp and a 400amp meters. It calls for 3 sets of 4 #350. But like I said only 2 conduits. How would I go about derating. I’m pretty sure I have too. Thanks. P.s. I cannot add another conduit considering the pad for the transformer is already poured.

As others have noted, running one run in conduit A and two parallel runs in conduit B is not kosher for _parallel_ conductors. I'm just going to look at the derating question.

You ask 'How would I go about derating' in this abnormal situation of trying to parallel conductors with different ampacity.

IMHO: If you have conductors of different ampacity in parallel, you would need to adjust the total ampacity to ensure that no single conductor gets overloaded. Again, for normal NEC parallel conductor installations this is simply prohibited.

Consider the OP's scenario where you have conductors of the same size but in different temperature conditions (the 2 sets in one conduit will run hotter than 1 set in one conduit). The impedance of the conductors will be very nearly equal, and thus the current will divide evenly between all three sets. This tells me that the ampacity of the single set ends up limited by the derating of the paired set to ensure that the conductors in the paired set don't carry too much current. So in this case simply figure out the derating factor for the 2 sets in one conduit, then apply that derating factor to all 3 sets.

As far as a practical solution to this situation, I like the approach of simply running two sets of 350 in each conduit, but this might hit problems on the ends with too many terminations for the various device terminals. If you are in one of the rare circumstances where the neutral counts as a CCC you might need to bump up to 2 sets of 400 in each conduit. If you try to replace 3 parallel 350s with 2 parallel larger conductors, you bump into the fact that larger conductors need greater conductor cross section per amp, and you need more copper cross section for the same current.

-Jon

 

[kwired]

As others have noted, running one run in conduit A and two parallel runs in conduit B is not kosher for _parallel_ conductors. I'm just going to look at the derating question.

You ask 'How would I go about derating' in this abnormal situation of trying to parallel conductors with different ampacity.

IMHO: If you have conductors of different ampacity in parallel, you would need to adjust the total ampacity to ensure that no single conductor gets overloaded. Again, for normal NEC parallel conductor installations this is simply prohibited.

Consider the OP's scenario where you have conductors of the same size but in different temperature conditions (the 2 sets in one conduit will run hotter than 1 set in one conduit). The impedance of the conductors will be very nearly equal, and thus the current will divide evenly between all three sets. This tells me that the ampacity of the single set ends up limited by the derating of the paired set to ensure that the conductors in the paired set don't carry too much current. So in this case simply figure out the derating factor for the 2 sets in one conduit, then apply that derating factor to all 3 sets.

As far as a practical solution to this situation, I like the approach of simply running two sets of 350 in each conduit, but this might hit problems on the ends with too many terminations for the various device terminals. If you are in one of the rare circumstances where the neutral counts as a CCC you might need to bump up to 2 sets of 400 in each conduit. If you try to replace 3 parallel 350s with 2 parallel larger conductors, you bump into the fact that larger conductors need greater conductor cross section per amp, and you need more copper cross section for the same current.

-Jon

He suggested 2 sets in each conduit, some of us just didn't read it that way for whatever reason.

 

[winnie]

He suggested 2 sets in each conduit, some of us just didn't read it that way for whatever reason.

Neither did I. I saw "4*4*350 mcm-2*4*350" and read that as somehow meaning 4 something in one conduit and 2 something in the other.

It was only on re-reading a couple of times that I realized it meant "4 sets of 4 conductors, 2 sets of 4 conductors in each of 2 conduits". But it took me a while. One of those perfectly correct things that looks perfectly wrong

-Jon

 

[Rriley88]

Let's assume that the person who came up with the 3 sets of 350's calculated the load. You didn't say copper or aluminum but let's assume aluminum. That's 3*250=750 amps. All you need to do is find the equivalent of 750 amps using only 2 sets of conductors. Parallel 750 aluminum would work (385*2=770 amps) or parallel 500 copper (380*2=760 amps). There are other possible combinations including using 4 smaller conductor sets, two sets in each raceway.

It is copper conductors

 

[Rriley88]

It is copper conductors

Here is a single line diagram of the service it’s it helps any

 

[electrofelon]

I see your choices as follows (and assuming you want close to or the full 700 amps which is likely overkill). CU conductors:
1. 1 set of 400 or 500 in each pipe
2. two sets of 3/0 in each pipe.
3. as was mentioned, serve some or all of the loads independently, i.e. not in parallel.

 

[Julius Right]

I agree, two sets of 2*4*2/0 -2*4*2/0 in each conduit-it will be enough.

 

[Julius Right]

Sorry.3/0 instead of 2/0- my mistake.

 

[infinity]

Here is a single line diagram of the service it’s it helps any

There should not be an EGC with the service conductors feeding the meters or the disconnect switches.