Parallel DC conductors - all positive in one conduit, all negative in another

[ChargedUp]

Do paralleled DC conductors feeding a single circuit with all positive in one conduit and all negative in another meet the requirements of 2017 NEC 300.3(B) and 300.3(B)(1), provided they are installed in accordance with 2017 NEC 310.10(H)?

In my experience, we typically see each set of positive and negative in same conduit, not all positives in one conduit and all negatives in another. For AC applications we rarely, if ever, run isolated phase. Regarding paralleled DC, is one practice better than another, or is it application dependent? Thanks

 

[wwhitney]

Looks like DC circuits are excluded from the requirements of 300.20, but I see no such exclusion on 300.3(B). So, is the conduit nonferrous?

If so, then underground it can meet 300.3(B)(1) Exception. Above ground, it can't meet 300.3(B)(1).

That leaves 300.3(B)(3), which I'm a bit unclear on. It would seem to me that it allows separated phase installation in all cases with non-ferrous conduit, which would render 300.3(B)(1) Exception moot. That interpretation is based on the term "sheath" meaning the conduit for conductors run in conduit. I don't see any other interpretation of 300.3(B)(3) that makes sense.

If the DC conductors are in ferrous conduit, I don't see how it can comply with 300.3(B), for better or worse. Unless I'm misinterpreting "sheath" in 300.3(B)(3), and it would apply to plastic insulated conductors in ferrous conduit, with the plastic insulation being the "sheath". In that case 300.20(B) is trivially satisfied for DC circuits. 300.20(A) would still prohibit separated phases in metallic conduits for AC circuits.

Cheers, Wayne

 

[texie]

Well, in transit systems DC +/- conductors are often in separate conduits that don't even follow the same path.

 

[masterinbama]

I've done large UPS systems( 725 KVA and larger) where we ran the positive and negative wiring in separate conduits and or cable trays.

Sent from my moto e (XT2052DL) using Tapatalk

 

[ChargedUp]

Thanks for the responses. In this case, the conduit should be PVC and it seems based on the posts above so far installing all positive and all negative in separate conduits would be compliant from an NEC standpoint. Please chime in if you disagree.

Otherwise, what about best practice: +/- in same conduit, or all positives in one, and all negatives in another?

 

[Besoeker3]

Are the DC conductors pure DC with zero AC ripple content?

 

[Carultch]

Thanks for the responses. In this case, the conduit should be PVC and it seems based on the posts above so far installing all positive and all negative in separate conduits would be compliant from an NEC standpoint. Please chime in if you disagree.

Otherwise, what about best practice: +/- in same conduit, or all positives in one, and all negatives in another?

Best practice is to populate each raceway in parallel, with all phases (or both polarities), neutral (if applicable), and ground, so that all conductors in each raceway can carry the full round trip path of the current, for that parallel set's fair share of the ampacity. Current should add up to zero in every conduit, so that the magnetic field surrounding it is reduced as much as possible.

Avoid isolating phases and polarities, because this magnetizes the surroundings. Steel conduit specifically, being built out of the namesake of ferromangetism, will amplify this problem. DC can be more forgiving to this problem, as it only produces a steady/constant magnetic field, rather than an alternating magnetic field that induces currents in the conduit as if the conduit were the secondary of a transformer (as happens with AC). But even still, it is best to avoid isolating DC polarities.

There is a section in the NEC on isolated phase installations, but it comes with a lot of context rules to follow, and it is very restrictive when you can do this.