EVSE and GFCI PROTECTION

[brycenesbitt]

Another person's research on EV and EVSE electrocutions:

Has anyone ever been electrocuted from an electric car? If so, what are the details?

George F. Rice's answer: I did the research, so you don’t have to (ahem). Many fun details of specific incidents follow the break. In short, I could find no record of any driver or passenger of a commercially manufactured modern electric vehicle ever being electrocuted. I found a lot of cases wh...

qr.ae

"Has anyone ever been electrocuted from an electric car? If so, what are the details?"

 

[retirede]

Another person's research on EV and EVSE electrocutions:

Has anyone ever been electrocuted from an electric car? If so, what are the details?

George F. Rice's answer: I did the research, so you don’t have to (ahem). Many fun details of specific incidents follow the break. In short, I could find no record of any driver or passenger of a commercially manufactured modern electric vehicle ever being electrocuted. I found a lot of cases wh...

qr.ae

"Has anyone ever been electrocuted from an electric car? If so, what are the details?"

Best line from the article. ”….though a lot of lawyers are writing hopeful blogs.”

 

[tortuga]

Correct. The issue is apparently high frequency noise from the AC to DC switching circuitry in the vehicle.

Thanks for confirming that, seems like alot of people are unclear on it, and that is probably where the NEC CMP's can surely wash their hands of it, its a on board vehicle charger issue.
5ma is the accepted safety level of leakage in wet areas, I don't think it matters if that 5ma is 50hz, 60hz, 400hz or DC but I am no expert.

 

[ActionDave]

Thanks for confirming that, seems like alot of people are unclear on it, and that is probably where the NEC CMP's can surely wash their hands of it, its a on board vehicle charger issue.
5ma is the accepted safety level of leakage in wet areas, I don't think it matters if that 5ma is 50hz, 60hz, 400hz or DC but I am no expert.

The 5mA mentality is the whole problem. It makes sense for a plug in your kitchen or outdoor receptacle but it makes no sense otherwise. We would all be safer with more ground fault sensing at a higher threshold for more types of circuits and equipment and not have the nuisance headaches.

 

[don_resqcapt19]

The 5mA mentality is the whole problem. It makes sense for a plug in your kitchen or outdoor receptacle but it makes no sense otherwise. We would all be safer with more ground fault sensing at a higher threshold for more types of circuits and equipment and not have the nuisance headaches.

Yet that was selected because about 95% of the adult population can let go with that amount of current flowing through their body. Higher currents can result in the person not being able to let go of the energized item.

 

[brycenesbitt]

Yet that was selected because about 95% of the adult population can let go with that amount of current flowing through their body. Higher currents can result in the person not being able to let go of the energized item.

Yet even better than 5mA is 0mA : isolated. That's what EVSE (and some hospital circuits) provide. And likely the reason nobody can find a record of injury to hang their code writing hat on. Extending what @ActionDave said think the whole problem through rather than getting stuck on 5mA. And use the data, Luke. Let's understand how people are actually getting injured and what the tradeoffs are.

 

[don_resqcapt19]

Yet even better than 5mA is 0mA : isolated. That's what EVSE (and some hospital circuits) provide. And likely the reason nobody can find a record of injury to hang their code writing hat on. Extending what @ActionDave said think the whole problem through rather than getting stuck on 5mA. And use the data, Luke. Let's understand how people are actually getting injured and what the tradeoffs are.

210.8(F) was never written with ESVEs in mind, and that or locating the charger outlet with 20' of a pool, are the only code rules that would require GFCI protection for hard wired ESVEs.

 

[wwhitney]

Yet even better than 5mA is 0mA : isolated. That's what EVSE (and some hospital circuits) provide.

No, EVSEs do not provide isolated power in the usual sense. That would require an isolation transformer with the secondary ungrounded and with ground fault monitors.

Cheers, Wayne

 

[brycenesbitt]

No, EVSEs do not provide isolated power in the usual sense.

Ok, fair enough. Then disconnected. Off until proven safe: a level of protection not afforded by even an NEC Article 100 5mA GFCI. A traditional dumb GFCI is constantly energized until it sees a fault. An EVSE relay is off until safety checks are passed.

 

[Fred B]

There are developing technologies that like the EVSE will not energize until the connection is made and a handshake is made between the the control unit and the utilization equipment. With these devices it is not necessary to have GFCI as no power will be present that can have user contact.
I'm still reading code as 'if it is direct wired no extra GFCI is required'. The EVSE will utilize either CCID5 or CCID20 with GMI per UL 2231-2 for personnel protection depending on engineering design that provides roughly the equivalent to GFCI.

Safety of Electric Vehicle Supply Equipment - IEEE Transportation Electrification Council

The IEEE Transportation Electrification Community coordinates broad and deep activities throughout the IEEE in the growing electrification revolution across transportation domains, including advances in electric and hybrid cars, more-electric ships and aircraft, rail systems, personal transport...

tec.ieee.org

 

[retirede]

There are developing technologies that like the EVSE will not energize until the connection is made and a handshake is made between the the control unit and the utilization equipment. With these devices it is not necessary to have GFCI as no power will be present that can have user contact.

That’s not a “developing technology”. Current EVSEs do not energize until the they sense connection to the vehicle.

There are cases where GFCI could still be valuable, such as a damaged charging cable.

 

[jaggedben]

210.8(F) was never written with ESVEs in mind, and that or locating the charger outlet with 20' of a pool, are the only code rules that would require GFCI protection for hard wired ESVEs.

So let's add an exception to 210.8 for EVSE that meet the requirements of 625. I gather the First Draft will be open for comments in July.

 

[don_resqcapt19]

So let's add an exception to 210.8 for EVSE that meet the requirements of 625. I gather the First Draft will be open for comments in July.

You can't add anything new in the comment stage.

However there was a PI that would add and exception in 210.8(F) for EVSEs. If that does not appear in the First Draft Report, comments could be made to bring it back for the Second Draft.
If it does appear in the First Draft Report, that does not mean that is will be in the 2026 code as comments, in addition to reviving a rejected PI, can delete an accepted PI. Then there is also the motion process at the 2025 NFPA annual meeting and after that there can be appeals to the NFPA Standards Council.

 

[jaggedben]

You can't add anything new in the comment stage.

However there was a PI that would add and exception in 210.8(F) for EVSEs. If that does not appear in the First Draft Report, comments could be made to bring it back for the Second Draft.
If it does appear in the First Draft Report, that does not mean that is will be in the 2026 code as comments, in addition to reviving a rejected PI, can delete an accepted PI. Then there is also the motion process at the 2025 NFPA annual meeting and after that there can be appeals to the NFPA Standards Council.

Thanks for the info. I'm satisfied that there will be an avenue for commenting on this general issue, and that it will be worthwhile for people here to let the CMP know their thoughts on it.

 

[ActionDave]

Yet that was selected because about 95% of the adult population can let go with that amount of currentou flowing through their body. Higher currents can result in the person not being able to let go of the energized item.

For areas like a kitchen or outdoor receptacle that makes sense. There are other places where the 5mA threshold is overkill. I'm probably not articulating my thoughts clearly but I'm using the same reasoning that allows GFPE level protection on heat tape circuits.

 

[Jpflex]

I believe the link I posted here in January answers your question.

Here is the link again

Notice the 6mA ground fault setting of type A, B and DD RCD.

Is the type A B and DD GFCI ratings referring to the level of mili amperes before tripping such as 4-5 mili amperes or does this also include environment or hazardous locations?

 

[ramsy]

does this also include environment or hazardous locations?

Good question for Rbalex

My guess is classified locations are not required in the IEC standards for RCD's, but may be commercially available from circuit breaker manufactures, such as Timelec.

 

[don_resqcapt19]

For areas like a kitchen or outdoor receptacle that makes sense. There are other places where the 5mA threshold is overkill. I'm probably not articulating my thoughts clearly but I'm using the same reasoning that allows GFPE level protection on heat tape circuits.

The protection of heat tape has nothing to do with protection of people...that is only an attempt to prevent fires.
If the let go current is correct, why would the location of where the person receives the shock make any difference?

 

[ActionDave]

The protection of heat tape has nothing to do with protection of people...that is only an attempt to prevent fires.
If the let go current is correct, why would the location of where the person receives the shock make any difference?

Because the ground fault is going sniff out the problem before anybody gets a shock.

Europeans use 30mA level protection almost everywhere.

 

[don_resqcapt19]

Because the ground fault is going sniff out the problem before anybody gets a shock.

Europeans use 30mA level protection almost everywhere.

The supply power to the charger is always on. The internal ground fault protection only looks for a ground fault when you connect to the vehicle and cannot provide protection for an internal equipment failure.

Their protection is based on preventing ventricular fibrillation and not tripping at the let go threshold. Two different concepts on how to protect people.

The NEC has moved in that direction with the requirements to use Special Purpose Ground Fault protection in some sections of Article 680. That is Class C or D protection, is only available as a stand alone device, not as a breaker, and requires an additional "ground check" conductor between the SPGFCI and the protected equipment. The trip level for those devices is 20mA.