Grounding in Enphase inverter solar PV

[harakumar]

I have some questions regarding grounding and I like to take your inputs.
System description:
1. The system is a 6.4 KW solar PV.
The system uses enphase inverter.
The ground for the enphase (Green wire) goes through the conduit all the way to the service entrance ground.
2. The panels are grounded using weeb clips , WEEB-DMC is used for equipment grounding for enphase inverter. WEEB bonding jumpers will be used to electrically connect spliced Unirac rail.
WEEB grounding Lug will be used to connect system to equipment ground.
3. Aluminium J-box is mounted on the rail and a conduit runs between the j-box and PV sub panel. The conduit will be connected to service entrance ground. The conduit runs in the attic of the building.
Goal:
1. Provide grounding for the equipment. The rails, metal parts etc..
2. Like to provide a safe path in case a power surge in rails etc due lightning near by.
3. Ensure over current protection devices operate properly.
Questions:
1. Can the rails be connected to a separate grounding electrode ? ( not connected to the service entrance ground). This gives a safe path in case of a lightning surge. The electrode will be close to where the panels are installed. This avoids the need for a separate equipment ground wire to service entrance for the equipment ground through the attic. The most important idea here regarding separate equipment ground is, if the metal parts picks up a lighting surge then it will go quickly to the separate ground and will not go through the attic to the service entrance ground. Shorter path and is safely routed to ground (separate ground).

2. Do I need a non-conducting spacer ( breaks the path) in the conduit so that the service ground and the separate equipment ground are separated( with out the non conducting spacer the conduit makes a continuous path from rail to the service entrance ground) .
a) Is separation safer approach as the surge will not have a path through the conduit ? But I think this is against NEC - am I correct?
Separate isolated ground is not allowed.
b) Is it a good thing to have the conduit conducting that is with out the non conducting spacer in the conduit? Using this method there will be a grounding electrode System as described in NEC 250.50 (2005). The exhibit 250.21 shows a metal fame connected to a grounding rod which is also connected to a concrete encased electrode. The conduit connected to service entrance ground will serve the path from the rack to the concrete encased electrode. The ground rod connected to the rack (the solar rack and metal parts) serves as another grounding method and they are both connected through the conduit. Need to ensure that the conductivity from the rack to j-box to conduit to service entrance ground exists.
Which approach is better?
I am thinking approach (b) is good. If a 4 AWG CU is used for rack to ground rod then this will work like a shunt path when there is a surge even though there is a path through the conduit (cannot guarantee it) or at least there are multiple paths to ground and a system of ground is better than just one service entrance ground.
c) A variation of method B is to keep the insulating spacer and remove the connectivity to ground through the conduit but have a separate wire run to the service entrance ground but around the building (not through the attic). This connects the ground rod (just for the rack) and the service entrance ground forming a grounding system.
Please give me your inputs. Which one is the best approach?
3. Like to know if the over current protection devices will operate OK with this arrangement.
Since the enphase ground comes to service entrance ground it should be OK - is this a correct statement?
Please give me you inputs.
Thank you

 

[jaggedben]

Questions:
1. Can the rails be connected to a separate grounding electrode ? ( not connected to the service entrance ground).

You can provide a separate grounding electrode for the solar array, but you still need a bonding jumper connecting it to the existing grounding system at the service panel.

This gives a safe path in case of a lightning surge. The electrode will be close to where the panels are installed. This avoids the need for a separate equipment ground wire to service entrance for the equipment ground through the attic.

No, you still need a separate equipment ground for the overcurrent protection to operate properly.

The most important idea here regarding separate equipment ground is, if the metal parts picks up a lighting surge then it will go quickly to the separate ground and will not go through the attic to the service entrance ground. Shorter path and is safely routed to ground (separate ground).

Good luck with lighting doing what you want it to. A direct lighting hit will destroy all of this anyway.

2. Do I need a non-conducting spacer ( breaks the path) in the conduit so that the service ground and the separate equipment ground are separated( with out the non conducting spacer the conduit makes a continuous path from rail to the service entrance ground) .

No. That does not make any sense and is not allowed by code. Again, you need continuous equipment grounding for overcurrent protection to operate.

a) Is separation safer approach as the surge will not have a path through the conduit ? But I think this is against NEC - am I correct? Separate isolated ground is not allowed.

Personally I don't believe it's predictably safer. And yes, it's against the NEC.

If you had a ground-mount solar system that was some distance from the house then there could be a meaningful debate as to whether you should have a ground wire provide a path for lighting to the house (although the NEC would still require it). However since you are talking about a rooftop system I don't think it makes anything predictably safer. The occupants are going to be in a lot of danger if lighting strikes the house, whether the grounding runs on the inside or the outside.

3. Like to know if the over current protection devices will operate OK with this (separate) arrangement.

I don't think they will.

Since the enphase ground comes to service entrance ground it should be OK - is this a correct statement?

No. Equipment grounding has to provide continuity back to the point of the neutral-ground bond in order to operate properly in the case of a hot-to-ground fault.

 

[ggunn]

No, you still need a separate equipment ground for the overcurrent protection to operate properly.

What has the EGC got to do with overcurrent protection? Ground fault protection I can see.

 

[jaggedben]

What has the EGC got to do with overcurrent protection? Ground fault protection I can see.

We're talking about the AC side (this is Enphase) and OCPD operation for a hot-to-ground fault

 

[harakumar]

Just a clarification.

Just a clarification.

What has the EGC got to do with overcurrent protection? Ground fault protection I can see.

Thank you.

There is about 50' distance from the solar array to the service panel. Even though the array is on
the roof it is still at a distance. A direct lightning is dangerous but lightning nearby can produce a surge due to electrical field and the goal is to protect damage from this so I am planning to add supplemental grounding electrode (like in ground install).

Just some clarifications:

1.
Is this OK to do:
The rack is connected to service entrance ground using #2 cu conductor.
The rack is also connected to a separate ground using #2 cu conductor.
So both the grounds are connected together through the rack.
Is this sufficient ? It is equivalent to one ground. The separate ground is a supplemental ground and gives an additional path t0 ground.
Exhibit 250.21 (NEC 2005) shows something similar.
Just trying to clarify that bonding directly is not needed if both the grounds are connected to the same conductor then it still means they are not isolated grounds.

2.
Second case:
The rack is connected to service entrance ground using #2 cu conductor.
The rack is also connected to a separate ground using #2 cu conductor.
These two are outside the building.
Conduit carrying enphase conductors are routed from j-box to the PV subpanel.
This conduit will be also be connected to service entrance ground (existing ground).
But since the conduit goes inside the building I like to put a nonconducting spacer before entering the building. Now one side of the conduit is connected to the rack which is grounded and conduit on the other side of the spacer is connected to ground at the service entrance. So everything is grounded but a surge due to lighting will not travel through the conduit. Is this OK to do?

Please reply.

 

[jaggedben]

1.
Is this OK to do:
The rack is connected to service entrance ground using #2 cu conductor.
The rack is also connected to a separate ground using #2 cu conductor.
So both the grounds are connected together through the rack.
Is this sufficient ? It is equivalent to one ground. The separate ground is a supplemental ground and gives an additional path t0 ground.
Exhibit 250.21 (NEC 2005) shows something similar.
Just trying to clarify that bonding directly is not needed if both the grounds are connected to the same conductor then it still means they are not isolated grounds.

I think that if the two #2 conductors are irreversibly spliced then you can consider them to be the required bonding jumper as well as GECs.

Are you on the 2005 code cycle? Precise interpretation of the code may depend on the exact code cycle.

2.
Second case:
The rack is connected to service entrance ground using #2 cu conductor.
The rack is also connected to a separate ground using #2 cu conductor.
These two are outside the building.

So far this is the same as above. Code does not care about inside vs. outside.

Conduit carrying enphase conductors are routed from j-box to the PV subpanel.
This conduit will be also be connected to service entrance ground (existing ground).
But since the conduit goes inside the building I like to put a nonconducting spacer before entering the building. Now one side of the conduit is connected to the rack which is grounded and conduit on the other side of the spacer is connected to ground at the service entrance. So everything is grounded but a surge due to lighting will not travel through the conduit. Is this OK to do?

Please reply.

Again, you have an equipment grounding requirement for the AC circuit that must be met. Either you must have an electrically continuous conduit, or a grounding wire in the conduit. In the 2011 code the grounding section of 690 explicitly prohibits using a grounding electrode conductor as an equipment grounding conductor, so you cannot rely on the #2 to clear a hot-to-ground fault current that might occur past the point of your nonconducting spacer. I believe your non-conducting spacer is therefore a code violation no matter how any of the other grounding is configured.

I'm not saying the nonconducting spacer is necessarily a dangerous idea, or that relying on the #2 GEC back to the service won't reliably clear a fault if installed properly. (It may or may not.) But I think it's a code violation.

If it is not so difficult to install the #2 wire to the service panel on the outside of the building, why not just run the conduit there as well?

 

[harakumar]

Thank you for your reply.