Bonding for Roof Top Solar Array

Status
Not open for further replies.

ESolar

Senior Member
Location
Eureka, CA Humboldt County
Occupation
Electrician/Contractor
Does the #6 Copper Wire EGC have to be continuous, or can a rack serve to bond two sections of the #6 Copper Wire EGC?

Aluminum Ironridge Rack #6 Copper EGC Aluminum Ironridge Rack #6 Copper EGC
=====================LUG-----------------------LUG=================LUG-------------------------[Panel]
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
With Ironridge you can do that (unless you run into that weirdo inspector who still thinks it's a GEC and not an EGC).

What the code says is that the racking must be 'identified as equipment grounding conductors'. What I think this means is that if the racking manufacturer instructions identify various hardware connections as providing bonding, you may rely on those connections instead of continuous copper. So you may bond the copper EGC to a rail with the Ironridge lug, and then if you want to put a bonding jumper way at the other end of that rail with more lugs, you may do so, including if you're using bonding splices in between to connect multiple pieces of rail. (Note that unless the distance between the first lug and the bonding jumper is more than about 15ft, you're wasting money on lugs. But not a code issue)

One thing you should not do in my opinion is rely on the solar panels to bond anything besides the rails they themselves are attached to. (see 690.48) i.e. do not run a bonding jumper, say, from the lower rail of one row of panels to the upper rail of a row below, if the only thing bonding those rails back to the main EGC is the UFOs holding down the upper row of panels.
 

ESolar

Senior Member
Location
Eureka, CA Humboldt County
Occupation
Electrician/Contractor
Reply

Reply

With Ironridge you can do that (unless you run into that weirdo inspector who still thinks it's a GEC and not an EGC).

...

One thing you should not do in my opinion is rely on the solar panels to bond anything besides the rails they themselves are attached to. (see 690.48) i.e. do not run a bonding jumper, say, from the lower rail of one row of panels to the upper rail of a row below, if the only thing bonding those rails back to the main EGC is the UFOs holding down the upper row of panels.

Thank you. That is what I thought, with the exception of the last part. While I do not intend to do such a thing (bond a rail by relying solely on a bond to that rail from another rail bonded only by panels, bonded to yet another rail bonded to the EGC, i.e., EGC --> Rail-1 --> Panel --> Rail-2 --> Rail-3), it is nevertheless interesting that one recommends against it because the bonds are supposed to be good. But I did notice that the ironridge documentation shows the wire EGC attaching to at least one of the two rails per row of panels (even where there are two rows, 4 rails). That is consistent with your advice. Again, thank you.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Thank you. That is what I thought, with the exception of the last part. While I do not intend to do such a thing (bond a rail by relying solely on a bond to that rail from another rail bonded only by panels, bonded to yet another rail bonded to the EGC, i.e., EGC --> Rail-1 --> Panel --> Rail-2 --> Rail-3), it is nevertheless interesting that one recommends against it because the bonds are supposed to be good. But I did notice that the ironridge documentation shows the wire EGC attaching to at least one of the two rails per row of panels (even where there are two rows, 4 rails). That is consistent with your advice. Again, thank you.

It's not because I don't trust the UFO bonds. It's because if you remove the panels from that row for servicing, it's not good if that unbonds the whole rest of the array. The 2014 code even has a section about putting in a bonding jumper when equipment is removed that provides bonding. But what service tech on this type of racking is going to think of that before removing panels?
 

Zee

Senior Member
Location
CA
It's not because I don't trust the UFO bonds. It's because if you remove the panels from that row for servicing, it's not good if that unbonds the whole rest of the array. The 2014 code even has a section about putting in a bonding jumper when equipment is removed that provides bonding. But what service tech on this type of racking is going to think of that before removing panels?

That is a good point, Jben.

OP, do you mean separate racks or rails? A rack is more of a complete structure vs a rail is just, well, a length of rail.
 

Zee

Senior Member
Location
CA
Haven't run #6 EGC in 15 years.

Just #10 under panels. (AWG equivalent to DC wire AWG required for DC)

If exposed to mechanical/physical damage then yeah #6.

So it is unclear if you are talking about separate racks with space between them where one may walk...or rails end to end under panels.......
 

pv_n00b

Senior Member
Location
CA, USA
The mounting system has to be listed to UL 2703 to be used as a replacement for an EGC conductor between components connected to the mounting system. Then you just need an EGC to connect the mounting system to the grounding system outside of the array or to connect separate array sections together.
 
Status
Not open for further replies.
Top