DC GEC directly to ground electrode

[souparvo]

Hi,

I'm on a dilemma. I want to have the DC ground electrode system to have an alternative in case the inverter is removed (690.49). As summary of the design, this is a flat rooftop system I have DC disconnects on every CB and they land on a pull box before they are routed from the roof to underground and to the inverter. For the GE system I have a ground plate that connects the inverter, main disconnect, existing building ground (is accessible) and a new ground rod.

So my suggestions are:

1. have a busbar on the pull box (NEMA 4) and bong the EGC from the CBs and have that busbar bonded to a GEC (#6 AWG Cu) and route it to the ground bar at ground level.

2. have a GEC going directly from the CBs (#6 AWG Cu) to the ground bar at ground level

My question is if I can actually place a busbar on a pull box and land my ECG there. I don't mind to continue the ground to the inverter and have it just irreversibly spliced and bonded to the pull box busbar instead of terminated there.

Thank you,

 

[jaggedben]

... My question is if I can actually place a busbar on a pull box and land my ECG there. ...

I don't see why not. It's what we do in integrated inverter discos all the time.

But I would leave enough of a service loop to be able to do the irreversible splice later if the AHJ requires it.

 

[souparvo]

I don't see why not. It's what we do in integrated inverter discos all the time.

But I would leave enough of a service loop to be able to do the irreversible splice later if the AHJ requires it.

What I'm not clear is if I need a disconnecting means in order to land EGC. The code only talks about a disconnecting means in order to land EG.

What I'm doing is landing the ECG on the pull box busbar and have a GEC going to the ground plate. If I continue the ECG to the inverter I would create a loop that could dangerously trap a fault current.

Thanky you for the response!

 

[jaggedben]

What I'm not clear is if I need a disconnecting means in order to land EGC. The code only talks about a disconnecting means in order to land EG.

I'm not sure where you're getting that from. The relevant code sections are 250.134 and 250.136 and they don't mention disconnecting means.

What I'm doing is landing the ECG on the pull box busbar and have a GEC going to the ground plate. If I continue the ECG to the inverter I would create a loop that could dangerously trap a fault current.

Continuing the EGC to the inverter in some form is required. Otherwise the ground fault detection in the inverter may not function, and that is probably the most dangerous mistake to make with a PV system. Under certain conditions, the EGC can be combined with a GEC or a bonding jumper between electrodes. But it has to go to the inverter.

From your descriptions, I'm not sure of the configuration you have in mind, or why you are worried about a ground loop. I guess I'm not clear on what you're trying to accomplish. You mentioned 690.49 above. That section, in my opinion, is somewhat misguided and should be removed or revised. But regardless of my opinion, I can't figure out what that section has to do with your question. Maybe you meant 690.48. But neither of those sections requires what you seem to have in mind, as they refer to temporary jumpers that would be installed if the inverter was removed.