Adding a parallel wire that is smaller

Status
Not open for further replies.
T

Tank11

Member
Location
CO
I have an installation with (2) parallel 500 MCM wires going into a building. There is room for another run but the only available wire is a 250 MCM.

If this third parallel wire is added, does it bring down the ampacity of the service to 3 times the rating of the 250 MCM, or does it add linearly to the parallel 500's?

Asking for a friend.
 
You cannot do that.

These are the requirements for parallel install

310.10(H)(2) Conductor and Installation Characteristics. The paralleled
conductors in each phase, polarity, neutral, grounded circuit
conductor, equipment grounding conductor, or equipment
bonding jumper shall comply with all of the following:
(1) Be the same length.
(2) Consist of the same conductor material.
(3) Be the same size in circular mil area.
(4) Have the same insulation type.
(5) Be terminated in the same manner
Click to expand...
 
Please pass on my regrets to your friend. What you described is not allowed by code. Reference NEC 310.10(G)(2). You can parallel conductors to get more ampacity in each phase, but all the phase A conductors must be the same as each other, and the phase B conductors must be the same as each other, and all the phase C conductors must be the same as each other.

NOTE: I see that Dennis beat me to the draw on this one. I also see he is using a different NEC edition than the one I have at my home office desk. I have the 2020. It is not yet accepted by WA State. But my 2017 is at my "real office" and I can't get there at present.
 
How many conductors _total_ are in the raceway?

As noted all conductors that make up a single phase must be the same, but phase A need not be the same as phase B. So you may be able to re-arrange the existing conductors and the pull new in to achieve your desired goal.

-Jon
 
winnie said:
How many conductors _total_ are in the raceway?

As noted all conductors that make up a single phase must be the same, but phase A need not be the same as phase B. So you may be able to re-arrange the existing conductors and the pull new in to achieve your desired goal.

-Jon
Click to expand...
I unddrstand what you are saying, but keep in mind that the only way that would work is if you used 3 x 250 on one phase, then 3 x 500 on the other two, with the circuit amps being limited to what the 3 x 250s can carry. With no adjusting for length, 3 x 250s would get you 765A, the 2 x 500s you already have got you 760A. Kind of pointless to get 5 extra amps...
 
The NEC does not allow you to parallel dissimilar conductor sizes. By parallel, it means both/all paths have the same origin device, and the same destination device.

This is true even if it would be code-compliant to scrap all but the smallest size, and have it carry the full load. If you are going to add a parallel conductor whether to curtail voltage drop or increase the ampacity, it has to be the same size as what it is paralleled with.

Dissimilar sizes & characteristics will mean that current will not divide uniformly among the paths in parallel, which means one path will systematically carry disproportionately more than its share of the current. To avoid complicating the matter and needing to calculate exactly how many amperes are carried on each path, the NEC ruled out dissimilar conductor sizes in parallel altogether.

Physics-wise, if you built dissimilar conductors in parallel despite not being code-compliant, current would divide in proportion to the conductance of each path. Conductance defined as 1/resistance. From the first order factors alone, conductance is proportional to the KCMIL, and inversely proportional to the length.
 
which means one path will systematically carry disproportionately more than its share of the current.

Please explain that - contrary to physics and circuit analysis it is, at least the way it is worded.

Now, the part about NEC dumbing it down so no calculation need, OK.
 
junkhound said:
which means one path will systematically carry disproportionately more than its share of the current.

Please explain that - contrary to physics and circuit analysis it is, at least the way it is worded.
Click to expand...


Put a 250 kcmil conductor in parallel with a 500kcmil.

The 250 has an ampacity of 255A. The 500, 380A.

The 500 has half the resistance of the 250, and so would carry 2/3 of the total current. If you had 630A flowing on the pair, it would divide as 420A on the 500 ans 210A on the 250.

Note the above ignores magnetic/inductive effects or the effects of differential temperature.

-Jon
 
winnie said:
Put a 250 kcmil conductor in parallel with a 500kcmil.

The 250 has an ampacity of 255A. The 500, 380A.

The 500 has half the resistance of the 250, and so would carry 2/3 of the total current. If you had 630A flowing on the pair, it would divide as 420A on the 500 ans 210A on the 250.

Note the above ignores magnetic/inductive effects or the effects of differential temperature.

-Jon
Click to expand...
All that, esp the last, is why they dumbed it down for us.
 
Carultch said:
The NEC does not allow you to parallel dissimilar conductor sizes. By parallel, it means both/all paths have the same origin device, and the same destination device.

This is true even if it would be code-compliant to scrap all but the smallest size, and have it carry the full load. If you are going to add a parallel conductor whether to curtail voltage drop or increase the ampacity, it has to be the same size as what it is paralleled with.

Dissimilar sizes & characteristics will mean that current will not divide uniformly among the paths in parallel, which means one path will systematically carry disproportionately more than its share of the current. To avoid complicating the matter and needing to calculate exactly how many amperes are carried on each path, the NEC ruled out dissimilar conductor sizes in parallel altogether.

Physics-wise, if you built dissimilar conductors in parallel despite not being code-compliant, current would divide in proportion to the conductance of each path. Conductance defined as 1/resistance. From the first order factors alone, conductance is proportional to the KCMIL, and inversely proportional to the length.
Click to expand...

Doesn't reactance and X/R also play a role?
 
If you had 630A flowing on the pair, it would divide as 420A on the 500 ans 210A on the 250.

Yes, as that IS proportionate, which is why I commented on the wording of your original post. Disproportionate only if considered in view of NEC RATED current.
 
mbrooke said:
Doesn't reactance and X/R also play a role?
Click to expand...

It would, but I limited my response to "first order factors alone" for simplicity. The effective resistance at the power factor is what really would ultimately matter. The values in Chapter 9 / Table 9, as opposed to Table 8. Resistance alone applies to DC, and is inversely proportional to kcmil regardless of the size. AC is different, because inductance factors and the skin effect apply. In general, it is close to inversely proportional to kcmil, but there is a deviation. Given the same length, the specific conductance in units of mhos per kcmil is less for larger conductors.

junkhound said:
If you had 630A flowing on the pair, it would divide as 420A on the 500 ans 210A on the 250.

Yes, as that IS proportionate, which is why I commented on the wording of your original post. Disproportionate only if considered in view of NEC RATED current.
Click to expand...

Disproportionate in terms of it not uniformly dividing. Yes it will be close to dividing proportionate to the KCMIL, but that isn't the point. The NEC's code making panel does not want to complicate the matter and require a calculation to substantiate whether a real life consequence occurs for a parallel assembly of multiple sizes. To keep it simple, the code just specifies to make each wire in parallel as identical as possible, with the expectation for current to uniformly divide.
 
Alright- though nit picking the values apply to single, not parallel sets. Though I do get your point and you are correct.
 
Status
Not open for further replies.
Top