Bonding of jbox attached to solar railing

[zman990]

I have a 4*4 deep 3/4" threaded hole cast aluminum bell box 3r rated that was connected to the solar rails that are bonded all together using the small clips that are provided in box to connect it to the solar rails using a stainless steel tek screw.
I failed inspection because I did not put a jumper from the the bell box and splice together with EGC coming into the box from the rails.
The conduit from the bell box is bonded at the inverter. And at the bell box since we used a rain tight emt connector and used channel locks to connect to box via threading.
I do not feel it is necessary to put a Bonding jumper for the box, and splice into egc from rails because box is bonded to rails even if there are splices in the the box.
The conductors in the conduit and bell box is also ungrounded since it is a TL inverter with ground fault protection and complies with 690.35.
It is bonded for lighting and since it already has ground fault protection that is insanely sensitive for dc site built into inverter.
So do I need a jumper inside box or are the aluminum 3r bell boxes sold at all electric distribution companies in the country bonded to rails?


Sent from my HTC One_M8 using Tapatalk

 

[ActionDave]

A metal box secured with a metal EMT fitting is bonded where I come from. There is some who will argue that since an EMT fitting does not have tapered threads it does not bond a bell box.

 

[jaggedben]

If you're splicing conductors inside a metal box then you need to bond the EGC to the box. It's no different from any other electrical installation. Can't remember the code section off the top of my head but I'm sure someone here can provide it. Maybe if you are not splicing anything in the box, and your conduit is properly bonded all the way back to the inverter, then it's okay not to have a bonding jumper to the box, but not too many people install a box if they're not splicing.

Whatever mounting screws you used to mount the box to the rail are not listed to bond the box to the rail, so I don't buy that part of your argument. For one thing, I've seen too many tek screws on solar arrays break, during and after install.

I run a solar installation company and I would insist to any of my crew that if they failed an inspection for this reason they should know better. Maybe it's not critical for safety given the TL inverter features, but it's certainly code.

 

[jaggedben]

... Can't remember the code section off the top of my head but I'm sure someone here can provide it. ...

Hey look, it's 250.148, which I was reminded of by this whole other thread going on right now.

 

[zman990]

Ok so since we used rain tight connector and bonded to the box with it and we bonded to the rails using a tek or machined screw and definitely used more than two threads. Then we also bonded the box to the rails.

It is good


ATTACH]13875[/ATTACH]

Sent from my HTC One_M8 using Tapatalk

 

[electrofelon]

View attachment 13874

Ok so since we used rain tight connector and bonded to the box with it and we bonded to the rails using a tek or machined screw and definitely used more than two threads. Then we also bonded the box to the rails.

It is good


ATTACH]13875[/ATTACH]

Sent from my HTC One_M8 using Tapatalk

A few comments: Note that a rain tight fitting has nothing to do with it. Also you havnt said whether there is a splice in the box. If there is not a splice and the box is properly bonded by an approved raceway, then you do not need a bonding jumper from the wire EGC to the box. As others have said, note that tek screws are not an approved bonding method. See 250.8. Regarding your attachment, I could not really make it out but it appears to be discussing grounding a metal box with a machine screw via a threaded hole. This is allowed per 250.8. I would not consider a Tek screw a "thread forming machine screw" or any other item in 250.8(A) Finally, note that 250.48 requires the bonding jumper if splices are made, even if the box is properly bonding by 37 other approved raceways

 

[zman990]

There are splices

Sent from my HTC One_M8 using Tapatalk

 

[zman990]

This still makes no sense to me that we are bonded just not when splices are present inside a box.
Also since it is ungrounded dc system why would there need to be a ground on dc side at all?


Sent from my HTC One_M8 using Tapatalk

 

[electrofelon]

This still makes no sense to me that we are bonded just not when splices are present inside a box.
Also since it is ungrounded dc system why would there need to be a ground on dc side at all?


Sent from my HTC One_M8 using Tapatalk

Honestly, I dont see the logic behind that requirement either but that is the way it is.

We still need grounding and bonding in ungrounded systems. Its for the same reasons as grounded systems (except for a few differences in what happens during a first fault):
1. to minimize touch potential between metallic parts
2. To allow the functionality of ground fault detectors during a first fault on an ungrounded system
3. to provide a low impedance path back to the source to trip an OCPD during a first fault on a grounded system or a second fault on an ungrounded system
3. To connect metal parts to earth to comply with 250.4


Also note that "ungrounded systems" in the context of transformerless inverters, are almost always not really ungrounded systems, they are still referenced to the utility or separately derived system they connect to which are nearly always grounded systems.

Here is some additional reading:

http://forums.mikeholt.com/showthread.php?t=173336
http://forums.mikeholt.com/showthread.php?t=171411

 

[jaggedben]

Also since it is ungrounded dc system why would there need to be a ground on dc side at all?

To sense dangerous ground faults. The ground-fault detection in the inverter will not operate properly without equipment bonding, no matter how sensitive it is.

 

[zman990]

I understand it won't work with equipment bonding and the inverter works like a champ go SE all the way.
Do they make a listed rec screw for ground if not I need to get on that.
I do feel it makes threads so maybe there needs to be one.
It is also stainless and not cheap.
This screw should never fail

Sent from my HTC One_M8 using Tapatalk

 

[jaggedben]

I understand it won't work with equipment bonding and the inverter works like a champ go SE all the way.
Do they make a listed rec screw for ground if not I need to get on that.
I do feel it makes threads so maybe there needs to be one.
It is also stainless and not cheap.
This screw should never fail

I've never seen a bell box that didn't come with a green screw, if that's what your asking.