PV Ground Fault Protection Question

[ForestGnome]

NEC 2023, 690.41 (B) states:

PV system dc circuits that exceed 30 volts or 8 amperes shall be provided with dc ground-fault protection meeting the requirements of 690.41(B)(1) and (B)(2) to reduce fire hazards.
Solidly grounded PV source circuits with not more than two modules in parallel and not on or in buildings shall be permitted without ground-fault protection.

I'm curious from a safety point of view why the NEC says it's fine to not have ground fault protection for ground-mounted string array. Why would adding an earth ground to single string PV system somehow make the system safe from ground faults? Is there something I'm missing?

Thanks!

 

[zanydroid]

NEC 2023, 690.41 (B) states:

I'm curious from a safety point of view why the NEC says it's fine to not have ground fault protection for ground-mounted string array. Why would adding an earth ground to single string PV system somehow make the system safe from ground faults? Is there something I'm missing?

Thanks!

Huh?

Solidly grounded is an electrical term, it’s not a common approach these days. Ungrounded is the standard. You will not be allowed to randomly ground an ungrounded MPPT and still be code compliant.

It reads to me as a very edge case restriction, so most modern systems will need GFP

I do wonder why this appears to limit current to 2P but implies you can go to as many volts as you want. I guess if there is a single ground fault on the + it will be brought down to a few volts above ground on both + and -.

That is not the case for ungrounded system. You will easily get 600V worst case across the system.

 

[ForestGnome]

My understanding was that "solidly grounded" meant having a connection directly to earth ground, which from what I have seen and read online seems to be standard for ground-mounted solar arrays (I see folks run an GEC from their arrays and mounting equipment to earth ground). Am I misunderstanding that term?

 

[zanydroid]

My understanding was that "solidly grounded" meant having a connection directly to earth ground, which from what I have seen and read online seems to be standard for ground-mounted solar arrays (I see folks run an GEC from their arrays and mounting equipment to earth ground). Am I misunderstanding that term?

That grounds the frames. The frames are, to one layer of abstraction, isolated from the DC+/DC- side. (They're not isolated, in the sense that they capacitatively couple to the solar cells).

Solidly grounded I take to mean one side of the electrical business part of the solar module (either the + or - depending on if positive or negative ground) are connected to EGC. The Solidly modifier is meant to exclude high impedance grounding (yet another kind of grounding, which i think actually comes up in modern solar in some kinds of ground fault detection), which isn't good for clearing faults.

Also GEC from array is not that common these days (since 2011?), it's jumped onto the EGC and allowed to be sized thusly.

Lots of Solar code archaeology going on in this thread :laugh:

 

[ForestGnome]

That grounds the frames. The frames are, to one layer of abstraction, isolated from the DC+/DC- side. (They're not isolated, in the sense that they capacitatively couple to the solar cells).

Solidly grounded I take to mean one side of the electrical business part of the solar module (either the + or - depending on if positive or negative ground) are connected to EGC.

Ah, thank you for the clarification, that makes sense. On a related note, is there an equivalent requirement for a GFD on the battery side? It seems to me that there isn't, but that seems strange to me.

 

[zanydroid]

Ah, thank you for the clarification, that makes sense. On a related note, is there an equivalent requirement for a GFD on the battery side? It seems to me that there isn't, but that seems strange to me.

I am not sure, but battery is a lot easier to limit than PV source circuits. There's a breaker or fuse right on the output of the one place outside the battery pack that's especially hot (+ terminal).

PV is nasty b/c it's just energized all the time, in a bunch of directions. It'll still be energized even if arc fault detector and ground fault detector trigger. At least with arc fault, if it's a series failure, the MPPT can stop drawing power which should address most of the issue. If it's a parallel arc fault, well that's not a good time, nothing the MPPT can do in a lot of cases other than yell at you to fix it.