Any issues with adding fuses to my DC Negative inside the combiner box?

[HX_Guy]

Also I had a chance to get up on the roof and inspect the combiner box situation a little closer and it seems the install will be a little easier and cleaner looking than I originally imagined.

The 6AWG white wire (going to the inverter) does indeed reach up to where I could put in another busbar similar to the PV+ busbar at the top. Plus, the leads for the PV- coming from the MC4 connectors reach all the way up to the fuse holders as well, I thought I was going to need to replace those with new MC4 connectors with longer wires. Also, 2 more fuse holders definitely fit on the DIN rail.

[HX_Guy]

Are you allowed to use that combiner box as a "raceway"?

In some installations you can't just run a wire through a box using it as a piece of conduit. Hopefully I am wrong on this one.

Thanks for clearing it up the "red" wire for me.

I...don't know. Would seem a little ridiculous for it to not be allowed...I mean you can have wires terminate there, but not pass through? What's the difference?

[SunEagle]

Are you talking about the red cable from in from the conduit and connection to the MC4 connector inside the box?

That is linking the PV- from one panel to the PV+ of of another panel. The reason it goes through the combiner is because one panel is on one end of the roof and the other is on another, so the cable goes through conduit in the attic and pops back up here at the combiner.

See red arrow:

Are you allowed to use that combiner box as a "raceway"?

In some installations you can't just run a wire through a box using it as a piece of conduit. Hopefully I am wrong on this one.

Thanks for clearing it up the "red" wire for me.

[bcroe]

Ignored by whom? Not by code... 2011 NEC 690.5(A) requires ground fault detection and interruption. 690.5(B)(1) says that the ungrounded conductors need to be disconnected in the event of a fault, and 690.5(B)(2) says that the inverter will cease to supply power to output circuits when this occurs.

More Edit:
In non-isolated transformers, the ground fault detection does not use a fuse to ground. There is more information here. Even though your Fronius has a ground connection through the fuse, because it isn't permanent, and no other bond exists, your system should be treated as ungrounded, and disconnects and OCPD are required on both + and -. The Fronius manual describes an installation as you've described, without a bond wire, as ungrounded.

One more edit... looking at the manual, it looks like the Fronius built-in fuses are on the DC+ only. That seems insufficient for an ungrounded system, and is maybe the source of my confusion understanding what you are trying to say. In any case, none of this seems especially applicable to the SolarEdge inverter, unless I am missing your point.

I am in agreement, and in compliance. Fronius must be too, because their internal DC disconnect
switch breaks both poles. An external disconnect should do the same? What I take here is
essentially all string PV systems would require breaking both poles; how many PV systems don't?

If the SolarEdge inverter isn't isolated, its disconnect would need to break both poles.

I know, there are other ways to detect a ground fault, than a fuse. Bruce

[sensij]

It is addressed here. Its pretty obvious if there is a ground fault in my array,
or just some leakage current over dozens of panels, there could be hundreds
of volts on either array lead. Working on the inverter DC wiring would not be
safe just because a single pole disconnect switch was opened. This all became
obvious when some array wiring had to be relocated, and the ground fuse blew.
Now there are disconnects for every array current carrying wire. This
issue seems to generally ignored. Bruce Roe

Ignored by whom? Not by code... 2011 NEC 690.5(A) requires ground fault detection and interruption. 690.5(B)(1) says that the ungrounded conductors need to be disconnected in the event of a fault, and 690.5(B)(2) says that the inverter will cease to supply power to output circuits when this occurs.

More Edit:
In non-isolated transformers, the ground fault detection does not use a fuse to ground. There is more information here. Even though your Fronius has a ground connection through the fuse, because it isn't permanent, and no other bond exists, your system should be treated as ungrounded, and disconnects and OCPD are required on both + and -. The Fronius manual describes an installation as you've described, without a bond wire, as ungrounded.

One more edit... looking at the manual, it looks like the Fronius built-in fuses are on the DC+ only. That seems insufficient for an ungrounded system, and is maybe the source of my confusion understanding what you are trying to say. In any case, none of this seems especially applicable to the SolarEdge inverter, unless I am missing your point.

[bcroe]

Heyco http://www.heyco.com/Solar_Power_Com...wer_Components

Those will keep the bugs out too, if everything else is really tight. Important in this
area on the ground. A little duct tape along some seams can make them really snug.

Bruce Roe

[HX_Guy]

The inside of the Hayco is a rubbery insert. You may be able to remove it and slit it to open it, and then when it is reinserted and tightened, it would hold it back together. I can't answer about the silicon.

Yes but I would still need to fit the connector though the round part that tightens it all together, no?

[Amy@altE]

The inside of the Hayco is a rubbery insert. You may be able to remove it and slit it to open it, and then when it is reinserted and tightened, it would hold it back together. I can't answer about the silicon.

[HX_Guy]

By the way what is that single red wire for?

Are you talking about the red cable from in from the conduit and connection to the MC4 connector inside the box?

That is linking the PV- from one panel to the PV+ of of another panel. The reason it goes through the combiner is because one panel is on one end of the roof and the other is on another, so the cable goes through conduit in the attic and pops back up here at the combiner.

See red arrow:

[HX_Guy]

Heyco http://www.heyco.com/Solar_Power_Com...wer_Components

That looks like a nice liquid tight connection BUUUT that would mean I would have to take apart the MC4 cable that residing inside the combiner box and I don't have the tool to reassemble it. So I ask again, can I put some silicon sealant on the outside?

[Amy@altE]

Heyco http://www.heyco.com/Solar_Power_Com...wer_Components

[SunEagle]

Remarking black insulation as white is OK, in fact, it is specifically allowed by 200.7(A)(3). Anyplace else in the circuit, the conductor will appear black, and be treated as unknown voltage.

Re-marking white insulation as black can only be done under a few conditions, the most common if which is if that conductor is part of cable, which is 200.7(C)(1). That is why it can be OK to use NM cable to wire a light switch, and just mark the white wire black. Anyplace else in the circuit where it is not marked, a visible white conductor not inside a cable would be seen as grounded, and may be misinterpreted as safe, or connected incorrectly. Of course, checking the voltage first is always a good practice, but code sometimes helps when good practice is inconsistent.

I agree with you. HX-Guy will probably have to replace it with a black insulated wire unless the Inspector says different.

I also agree with you that it doesn't matter what the color of the wire you first need to check with a meter to see if it is alive or dead.

[bcroe]

Bruce, You might to read this to get a better understanding of how modern transformerless (AKA, non-isolated, aka ungrounded) inverters work. Yes, it is required that any disconnect and OCPD on the DC side is on both the DC+ and DC- in those circuits.

Thanks for that information. I seem to agree with this. As for inverters, the issue
is input-output ISOLATION, and it would be less confusing if everybody used
the same and consistent description, instead of using the round about and NOT
conclusive term "transformerless".

"Inverters that include an isolation-transformer as part of the listed product are
specifically designed and tested for use with grounded PV systems and are
often referred to as isolated inverters or transformer-isolated inverters."

I see the SolarEdge-Optoimizer system is not isolated. So a dual pole disconnect
switch would be required. That also explains why a specific inverter input voltage
must be maintained. "SolarEdge has developed a unique utility-interactive
PV system that consists of module-level dc-to-dc power optimizers coupled with
proprietary non-isolated inverters."

I note, they mention floating vs bonded conductors. Here the Fronius neg is
NEITHER. The array negative isn't FLOATING, because its connected by a
very low impedance (fuse) to ground. It isn't BONDED, because the fuse can
be blown open and in fact is expected to do so for a fault. This issue is not
addressed in the document.

It is addressed here. Its pretty obvious if there is a ground fault in my array,
or just some leakage current over dozens of panels, there could be hundreds
of volts on either array lead. Working on the inverter DC wiring would not be
safe just because a single pole disconnect switch was opened. This all became
obvious when some array wiring had to be relocated, and the ground fuse blew.
Now there are disconnects for every array current carrying wire. This
issue seems to generally ignored. Bruce Roe

[sensij]

I am not sure if NEC 200.7 (C) can be used for this situation or not. It can sometimes get a little fuzzy when you see one thing and then read about it in the code as not being acceptable.

Attached is a combiner box used in a solar pv array. The DC wires coming from the panels are white and yellow insulated and the wires going to the inverter are black insulated with white and yellow tape. It (along with 11 others) passed inspection.

Remarking black insulation as white is OK, in fact, it is specifically allowed by 200.7(A)(3). Anyplace else in the circuit, the conductor will appear black, and be treated as unknown voltage.

Re-marking white insulation as black can only be done under a few conditions, the most common if which is if that conductor is part of cable, which is 200.7(C)(1). That is why it can be OK to use NM cable to wire a light switch, and just mark the white wire black. Anyplace else in the circuit where it is not marked, a visible white conductor not inside a cable would be seen as grounded, and may be misinterpreted as safe, or connected incorrectly. Of course, checking the voltage first is always a good practice, but code sometimes helps when good practice is inconsistent.

[SunEagle]

Hey guys, real quick on a different note, what do you recommend for making the left side conduit bushing water tight? My installer just used what looked like a anti-short bushing and today we got some rain, and sure enough, I checked the combiner box and there was water inside (I shut off the system this morning before it started to rain as a pre-caution).

Could I just put some silicon caulking on the outside of the bushing to seal it tight? Would that be ok?

You might have to replace it with a compression fitting for cables. Like this Hubbell fitting. http://www.cesco.com/b2c/product/120...FUQV7Aodyk0AIQ

By the way what is that single red wire for?