Am looking for some electrical code help

[zanydroid]

What about

300.20 Induced Currents in Ferrous Metal Enclosures or Ferrous Metal Raceways

(Though it doesn’t explicitly say the current needs to be balanced). If you run the conductors in ferrous conduit, then it needs to be either L-L-N , or if L-L the conductors need to come back on the same return path. In this latter case there are 4 current carrying conductors so derating would be needed.

If you run the conductors in non-ferrous conduit you are fine from conduit angle, but when the imbalanced current enters a ferrous box considerations need to be made. For instance if you have a ferrous pull point or splice point. Or if you originate and return to the same ferrous box. In the latter case you would need to either make them come back in the same hole (maybe by using a non ferrous JB or conduit body to regroup them) or make appropriate cuts in the ferrous box between the holes to block the eddy currents

For 300.3, switched legs are allowed to exclude neutral (those have balanced current) but are required to have lines in same path even if non ferrous.

 

[zanydroid]

If you are running into conduit fill problems that implies you also need to make sure the current carrying conductors are applied, if those haven’t been applied yet your conduit fill would probably go over after you upsize the conductors.

The eddy currents would be locally induced in the ferrous enclosure or conduit along the whole path so I don’t think distance matters.

 

[zanydroid]

Regarding conduit fill and derating. Have you considered using individual aluminum SE, SEU, or SER cable? Those are non-ferrous and pretty inexpensive. They are available with 2 and 3 insulated conductor variants.

 

[pvgirl]

On conduit fill remember this note form the NEC:

Where conduit or tubing nipples having a maximum length not to exceed 600 mm (24 in.) are installed between boxes, cabinets, and similar enclosures, the nipples shall be permitted to be filled to 60 percent of their total cross-sectional area, and 310.15(C)(1) adjustment factors need not apply to this condition.

 

[LBen]

Hey, all good input! I have to run off to an electrical supply store 70 miles away. When I get back, I'll install the AC disconnect and the meter socket. Then I'll take pictures of the outside and inside configurations and post them along with the final SLD that the state approved. That will hopefully give your folks a better idea of what I've been dealing with.

 

[Hedges]

What I keep getting hung up on is that I've got two AC paths to the end user load (main breaker panel). For one path, the grid goes through the xfer switch and then to the breaker panel. The other path has the grid going to the Sol-Ark whose load output goes to the xfer switch and then to the breaker panel. Both L1/L2 paths use the same neutral.

If current flows in L1, L2, and any imbalance in N, I would like all three to be within a single cable or conduit. I don't like having the N current split off to a different conduit.

I have conducted some experiments at 60 Hz and modest current around 10A where ferrous conduit separated return current, and I could not measure any extra voltage drop. I did this after hearing tales of steel parts glowing red hot, and I wanted to determine the effect of routing a PE wire in ferrous conduit. Maybe the tales of heating involved something with lots of high frequency harmonics?

But I think code requires the circuit to all be in a single conduit.

I would not mind if neutral split into two parallel paths where circuit was split by transfer switch, and connected back together (formed a loop) at the load. So long as each path can carry the full current. If inductance comes into play, the current will choose the closer neutral wire. So I would be inclined to make a neutral loop, also a ground loop.

Another approach is to branch out from transfer switch to SolArk, then output of SolArk goes back through same conduit to transfer switch before continuing on to load. My interlocked breaker transfer switch is wired exactly like that. But it does mean you have 4 current carrying conductors, so derating applies. And 6 total wires, including one N and one (smaller) G.

Where conduit or tubing nipples having a maximum length not to exceed 600 mm (24 in.) are installed between boxes, cabinets, and similar enclosures, the nipples shall be permitted to be filled to 60 percent of their total cross-sectional area, and 310.15(C)(1) adjustment factors need not apply to this condition.

I noticed this allows heat to be generated in that 24 in., but doesn't address where it will be dissipated. If the wire continued 10' in each direction outside the overstuffed nipple, there would be somewhere for heat to flow. If it landed on a lug of transfer switch just outside nipple at one end, and lug of SolArk at the other, it would drive those lugs to elevated temperature. SolArk's output has breakers, so those lugs would push the breakers closer to tripping. (something SolArk itself doesn't deliver enough current for, but pass-through from grid could.)

 

[zanydroid]

60% is not that far off from the 53% allowed for arbitrary distances for single conductor/cable, and I think the air flow would be more consistently sane in a nipple. As well the same airflow for a bounded amount of total heat inside the conduit.

Regarding making cuts vs trying to regroup the conductors to cancel eddy currents on the edge of a ferrous box. Cuts are probably preferred because you most likely need a monster size box to be code compliant for the conductor sizes I think you have.

 

[Hedges]

I was planning on running a neutral out to the transfer switch, but now wonder if that is just a waste of wire?

There's no need to have a neutral at the transfer switch. This isn't correct

Wouldn’t there be imbalanced current along that section of the circuit in the absence of a neutral?

I had checked with one of the inspectors and he agreed that I only need a two pole transfer switch (I don't need to switch the neutral). So I think the neutral is not used at the transfer switch. I think in one of @Hedges replies he mentioned that he had several neutral branches on his setup.

In most cases we don't switch the neutral. Neutral would not connect to the transfer switch.

If L1/L2 run through transfer switch and out a different conduit, I would want N to follow the same path, unless L1/L2 current were always identical (240V only load.) If this is a transfer switch with 3 conduits attached, N would splice so if branches out. Sometimes switch has a lug with optional green screw to bond case. If used for N, would need another lug for ground.

I have a safety switch (not transfer) hanging off a Tee. L1 and L2 go through it, and output L1 and L2 return through same conduit. Although L1 and L2 are not balanced, L1 in is balanced to L1 out, and L2 in is balanced to L2 out, so N doesn't go to/through the switch, just runs past through Tee to load.

I happened to hang surge arrestor off this switch, but it clamps L1/L2 to G, not N (which is bonded to G upstream.)

 

[zanydroid]

Another consideration for pulling neutral is that it can make future reconfiguration easier (sort of like how we nowadays pull neutral to every lighting switch box).

Pretty complicated to analyze though whether it's worth it in the general case.

 

[pvgirl]

Seems like if you had all 3 L1/L2 wires run together to the transfer switch, the currents would always balance between the 2 active L1/L2 conductors, and the neutral wouldn't be needed to run to the transfer switch.

 

[LBen]

Another long 12 hour day. I'm supposed to be retired. I'm at burn out stage, so will review all comments sometime tomorrow.

I'm attaching two pictures showing my current hardware configuration, both inside and outside. I'm also including the SLD I believe the state accepted. It pretty much is in line with one of Sol-Ark's line diagrams.

 

[zanydroid]

Is the bypass switch -> wireway 2 feet? So that you qualify for the conduit fill? Looks like it. I think you should be able to do L1/L2 out and back and satisfy what we discussed above with balancing the current. And then where the unbalanced L1/L2 comes up into the subpanel from the bottom you can fix that by running the neutral through the same hole.

Is your feeder to the pedestal coming from the bottom (visible from inside picture).

Your SLD has the PV conductors in PVC "inside" the garage but it comes up from the ground up to the wireway, so that addresses one concern I had with that. That white conductor is probably not OK for ungrounded PV system, and the inspector may not like seeing the PV go through the same hole as the AC / without separation inside the boxes & wireways.

I'm pretty confused by what the production meter is going to achieve in this SLD when used with a hybrid. What is the POCO going to use the numbers for? In case the numbers end up being bonkers (IE useless) because of how hybrids work (IE under most configs you can't see how much solar is actually generated with an external device).

What are those polaris connectors for? I don't see it in the SLD

 

[LBen]

I need to get going after a rather restless night worried about neutrals and reactance. Need to visit airplane this morning to verify a hydraulic fix with a test flight and to see if I can still land after not flying for two plus weeks.

Is the bypass switch -> wireway 2 feet? So that you qualify for the conduit fill? Looks like it. I think you should be able to do L1/L2 out and back and satisfy what we discussed above with balancing the current. And then where the unbalanced L1/L2 comes up into the subpanel from the bottom you can fix that by running the neutral through the same hole.

It is 24.5" from box to box. Don't know if inspector would count the connector length on either end. But I'm willing to let it go for now to see if he complains. Should have bought a 48" gutter instead of the 36".

Is your feeder to the pedestal coming from the bottom (visible from inside picture).

Yes

Your SLD has the PV conductors in PVC "inside" the garage but it comes up from the ground up to the wireway, so that addresses one concern I had with that. That white conductor is probably not OK for ungrounded PV system, and the inspector may not like seeing the PV go through the same hole as the AC / without separation inside the boxes & wireways.

Colorado is now under NEC 2023 and I think the code is more lenient with allowing PV in the same conduit with AC. I need to check on that. By "white conductor" are you referring to the feeder/service neutral? That is just a temporary connection until I get all this neutral stuff figured out.

I'm pretty confused by what the production meter is going to achieve in this SLD when used with a hybrid. What is the POCO going to use the numbers for? In case the numbers end up being bonkers (IE useless) because of how hybrids work (IE under most configs you can't see how much solar is actually generated with an external device)

Same here. I had several postings where I was questioning this. I had several discussions with the PoCo solar rep about why do I need a PM for a hybrid system? I told him that I didn't see how the PM was going to register anything different than the net meter. The net meter is bi-directional while the PM is uni-directional. He claims that they will have different readings. I also pointed out that the PM will not see house consumption. He said that they are only interested in what my system would push back into their system. Ok, to he?? with it, I'll spend the 3-400 bucks to put in a useless PM just to make them happy on paper. I suspect they have their meter reading system set up configured to pull numbers from two meters if the address is listed as having solar.

What are those polaris connectors for? I don't see it in the SLD

I have to split the incoming service/feeder to two locations; (1) xfer switch so that I can bypass the solar if I ever need to, (2) the Sol-Ark. The SLD shows the branching, but I didn't call out that I was using Polaris connectors.

 

[zanydroid]

By "white conductor" are you referring to the feeder/service neutral?

No I was referring to the small white one coming out of the PVC conduit, in the bundle of PV conductors

 

[zanydroid]

I suspect they have their meter reading system set up configured to pull numbers from two meters if the address is listed as having solar.

Ok if they just need to see what is being pushed back then sure…

I thought a production meter is , for production , but whatever. I’d recommend looking at the first few bills closely to make sure they use the reading in a legit way.

 

[LBen]

No I was referring to the small white one coming out of the PVC conduit, in the bundle of PV conductors

There is a red and black in the bundle also. If the inspector complains about a white PV wire, I'll have to point out that they all carry DC and that I purposely picked 7 different colors so that I can easily track string pairs.

 

[LBen]

I’d recommend looking at the first few bills closely

Agreed. The PoCo only pays 1.6 cents per kWh pushed back into their system. I will be comparing both meter readings and checking against the Sol-Ark. There is a third party app (I can't remember the name) that I may also use to monitor things.

 

[LBen]

For 300.3, switched legs are allowed to exclude neutral (those have balanced current) but are required to have lines in same path even if non ferrous.

Could you point to specific part of 300.3? I've read 300.3 several times and can't locate that specific reference.

 

[zanydroid]

Could you point to specific part of 300.3? I've read 300.3 several times and can't locate that specific reference.

It doesn’t mention switched legs (for lights) specifically. That is a standard and code compliant way to wire lights though. All non neutral conductors in circuit are run together, switched legs are not done with single conductor cables even with non ferrous boxes and conduit.

 

[zanydroid]

There is a red and black in the bundle also. If the inspector complains about a white PV wire, I'll have to point out that they all carry DC and that I purposely picked 7 different colors so that I can easily track string pairs.

The issue is that white is a grounded DC color and you have an ungrounded system.

This is somewhat ambiguous in NEC. Some cities explicitly have rules saying not to use white.

Frankly them enforcing it probably makes zero practical safety difference since you have it bundled up with a ton of high voltage PV lines so it should be clear.

It is 24.5" from box to box. Don't know if inspector would count the connector length on either end. But I'm willing to let it go for now to see if he complains. Should have bought a 48" gutter instead of the 36".

I’d leave it as is, most inspectors aren’t going to bust out a ruler. If they do you might have to resize some breakers due to derating. Or not, if the derated value is between sizes you can round up.