2026 NEC draft - EV and GFCI problems incoming?

straight from the source, they have expanded GFCI to hardwired single phase EVSE now too:

https://docinfofiles.nfpa.org/files/AboutTheCodes/70/70_A2025_NEC_P12_SD_BallotFinal.pdf page 84



There was very minimal justification from the public inputs on first draft

https://docinfofiles.nfpa.org/files/AboutTheCodes/70/70_A2025_NEC_P12_FD_PIResponses.pdf page 224





Second draft had several comments trying to remove it, but the committee statement was basically just ‘we expanded 625.54 for GFCI for receptacles and outlets.
https://docinfofiles.nfpa.org/files/AboutTheCodes/70/70_A2025_NEC_P12_SD_PCResponses.pdf page 146



this panel doesn’t have a lot of breaker MFG representation, I’m surprised such a big change passed through with so little resistance. Interestingly, there were a few long dissenting opinions on the final ballot, one from the chair.
https://docinfofiles.nfpa.org/files/AboutTheCodes/70/70_A2025_NEC_P12_SD_BallotFinal.pdf page 84
 
From what I recall it was forum member Bill Snyder here that made that proposal, Don mentioned it here: https://forums.mikeholt.com/threads/nec-or-ul.2582682/post-2940261
I hope your state can amend the code to remove this nonsense requirement. We have discussed on here many times how EVSE ground fault protection works:

EVSE and GFCI PROTECTION

Am I reading 625.54 correctly, if EVSE is plug and cord connected it must be GFCI protected, but if direct connected only if the manufacture specifies in the instructions? So in direct connecting an outlet for an EVSE it would be possible to skip the GFCI, UNLESS CALLED FOR BY THE MANUFACTURE...
forums.mikeholt.com
 
gene6 said:
From what I recall it was forum member Bill Snyder here that made that proposal, Don mentioned it here: https://forums.mikeholt.com/threads/nec-or-ul.2582682/post-2940261
I hope your state can amend the code to remove this nonsense requirement. We have discussed on here many times how EVSE ground fault protection works:

EVSE and GFCI PROTECTION

Am I reading 625.54 correctly, if EVSE is plug and cord connected it must be GFCI protected, but if direct connected only if the manufacture specifies in the instructions? So in direct connecting an outlet for an EVSE it would be possible to skip the GFCI, UNLESS CALLED FOR BY THE MANUFACTURE...
forums.mikeholt.com
Click to expand...

Correct, page 224 on this link shows it was submitted by William Synder from RCC solutions

IF he's on this forum I'd love to hear the rational since all that is written on the PI is " The problem is in the interpretation that EVSE when hardwired is required to have class A GFCI protection per 210.8(A)-(F)"
 
Yeah, from the thread gene linked, Bill is on this forum. GFI makes sense when it's an outlet and cord type setup, as the cord is likely laying on the floor of a potentially wet garage, but hardwired makes no sense to me since the charger itself has GFI built in as part of the design.
 
retirede said:
That’s not a valid reason. Look at 625.17(A)(3). The cord cannot lay on the floor.
Click to expand...
Tell that to the SecOps weenie who just pulled in his driveway after sitting in traffic for two hours following a 10 hour work day rebuilding customers' Ivanti firewalls. I guarantee you he has no clue that 625 even exists, nor cares enough to carefully route his power cord to meet the requirements ;)
 
Is there a allowance for GFI that is integral to the EV charger unit?

I can't tell by reading it.

For example, if Tesla's wall unit installed a GFI breaker integral to the housing that could be reset remotely, like via the app? That might be a solid fix and a useful upgrade to the charger.
 
From what I saw (this coming from the Motortrend article I linked in the OP), it looks like extant wall chargers like Tesla's already include GFI testing and protection, they just trigger at a higher level (about 15-20mA) than a breaker allows (5mA), yet still lower than the EU trip point (30mA).
 
ModbusMan said:
since the charger itself has GFI built in as part of the design.
Click to expand...
It does not have GFCI (5ma trip level). It likely has CCID-20 (20 ma trip level).

It seems to me that this is a case of one group of individuals (designers and writers of various EVSE standards) spending decades looking a safety issue and deciding that 5ma trip may not perform well, but that sufficient safety can be achieved with a higher trip level as long as the standards require additional measures, such as EGC assurance. And then another group of individuals (the CMP in charge of Article 625) overruling the judgement of the subject area experts, likely with comparatively minimal consideration.

Is there much chance of this change being reversed "on the floor" at the final vote? If not, I expect we will see EVSEs come out that implement CCID-5, which is part of the standard, possibly with a couple changes to comply with the GFCI standard. And we'll see if the original designers were correct in their judgement that it would lead to excessive nuisance tripping.

Cheers, Wayne
 
wwhitney said:
It does not have GFCI (5ma trip level). It likely has CCID-20 (20 ma trip level).
Click to expand...
I'd say its more about trip time than trip level, UL 943 is the only Class A standard that requires the 5-7ma trip level and UL 943 is not mandated by the NEC by reference.
In the UK ( BS-7671 ) and EU (IEC) equivalent standards for 'Class A ground fault protection' have been long adopted world wide and allow for 30ma of protection. They call them Residual Current Devices (RCD).
In theory a GFCI device could meet the NEC definition of a GFCI and not have that 5-7ma trip level.
Here they have been playing catchup with RCD's creating stuff like CCID and SPGFCI with different mA trip levels, using the same time to trip formula as a RCD and GFCI does. I would expect now that this has passed we'll see manufacturers adopt a RCD as a Class A GFCI, thus solving all the 240V nuisance trip problems (EVSE, Fridges, HVAC etc).
 
Last edited:
tortuga said:
a device could meet the NEC definition of a GFCI and not have that 5-7ma trip level.
Click to expand...
I don't see how that's possible, given the NEC definition and the Informational Note from the 2023 NEC, excerpted below.

Cheers, Wayne

Ground-Fault Circuit Interrupter (GFCI). A device intended for the protection of personnel that functions to de-energize a circuit or portion thereof within an established period of time when a ground-fault current exceeds the values established for a Class A device. (CMP—2)

Informational Note: See UL 943, Standard for Ground-Fault Circuit Interrupters, for further information. Class A ground-fault circuit interrupters trip when the ground-fault current is 6 mA or higher and do not trip when the ground-fault current is less than 4 mA.
 
wwhitney said:
I don't see how that's possible, given the NEC definition and the Informational Note from the 2023 NEC,
Click to expand...
Informational notes are not enforceable.
There is nothing in the NEC that mandates following UL, just that its Class A protection.
The rest of the world looked at UL 943 and passed on it in favor of the RCD standards.
There is nothing preventing CSA or ETL from adopting a RCD standard that provides Class A protection.
 
tortuga said:
Informational notes are not enforceable.
Click to expand...
No, but it's telling you that when the definition says "Class A" it means UL 943 Class A. Just like Appendix A does. I don't see how you can use a different standard when the informational note directs you to a particular standard.

Cheers, Wayne
 
tortuga said:
There is nothing in the NEC that mandates following UL, just that its Class A protection.
Click to expand...
So next you're going to tell me that while 240.7 requires circuit breakers to be listed, nothing requires them to be listed to UL 489? UL 489 incorporates the requirements for GFCI circuit breakers, and it references UL 943.

Note also that the definition of "listed" certainly allows the AHJ to judge which standard the item should be listed to, and any reasonable US AHJ is going to choose UL 489.

Cheers, Wayne
 
wwhitney said:
No, but it's telling you that when the definition says "Class A" it means UL 943 Class A. Just like Appendix A does. I don't see how you can use a different standard when the informational note directs you to a particular standard.

Cheers, Wayne
Click to expand...

See 90.5(C)
(C) Explanatory Material. Explanatory material, such as references to other standards, references to related sections of this
Code, or information related to a Code rule, is included in this Code in the form of informational notes.
Such notes are informational only and are not enforceable as requirements of this Code.
Click to expand...
 
tortuga said:
See 90.5(C)
Click to expand...
No, the informational note's reference to UL 943 is not incorporating UL 943 into the NEC as enforceable language. Yes, it is telling you that the enforceable terminology "Class A" in the GFCI definition means Class A as per UL 943. The informational note is material that elucidates the enforceable material, rather than providing additional enforceable text.

See also my previous post. Interestingly, there is not general requirement I see that GFCIs be listed, but certainly for GFCI circuit breakers, 240.7 covers it.

Cheers, Wayne
 
wwhitney said:
Yes, it is telling you that the enforceable terminology "Class A" in the GFCI definition means Class A as per UL 943. The informational note is material that elucidates the enforceable material, rather than
Click to expand...
Nothing in a informational note would be admissible as enforceable in court. The term Class A is probably also unenforceable.
All the major manufacturers now merge the two standards (UL and IEC) so there are many breakers that meet UL and IEC 60891 or 60947.
The NEC is used in a handful of other countries (AHJ) alongside IEC standards now, look at how UL 508 got dissolved into IEC.
 
You must log in or register to reply here.
Top