Suitability of NEMA 14-50 for EV charging -- Child Safety & Thermal Cutoff

brycenesbitt

Senior Member
Location
United States
So I know the NEC does not directly regulate what gets plugged into what outlets, but that said there's a connection.
It seems like NEMA 14-50, borrowed from range and dryer circuits, is becoming the standard for EV chargers. Yet there are some issues:
  1. The available plugs & connectors are rated for a low number of cycles, yet some EV owners travel with their charger.
  2. It’s possible to get electrocuted when plugging in or removing, due to lack of shrouded pins.
  3. The neutral adds 30% or so to copper costs, yet is not needed for EVSE equipment.
  4. The presence of a neutral opens up the chance someone will buy an unfused 14-50 to 5-20 outlet adapter.
  5. When the cord is unplugged, children can access the pins.
  6. There's no reliable make before break for the power pins.
While I have not seen a melted 14-50 in person, pictures of melted ones are all over the Internet:

What do people think of 14-50 as the connector for this purpose long term?

Would there be appetite within the NEC code committees to require a child lock in service cover for this install, especially when installed at child height?
To require that "smart" connected devices monitor the temperature of the plug?
Has there been any NEC discussion of IEC 60309-1 and IEC 60309-2 connectors for this purpose?
 

Attachments

  • NEMA EV Charger Melt.jpg
    NEMA EV Charger Melt.jpg
    283 KB · Views: 25
  • IEC 60309 Round Sleeve.JPG
    IEC 60309 Round Sleeve.JPG
    84.7 KB · Views: 25
  • IEC 60309 Example.jpg
    IEC 60309 Example.jpg
    8.8 KB · Views: 26
  • NEMA Melted Outlet Box.jpg
    NEMA Melted Outlet Box.jpg
    487.9 KB · Views: 29
  • NEMA Amazon Illegal 14-50 to 5-20 adapter.jpg
    NEMA Amazon Illegal 14-50 to 5-20 adapter.jpg
    49.9 KB · Views: 27

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Not directly answering your questions but there was a thread a little while back about 14-50s that melt vs. better quality ones that don't.

I think you're probably right that a better standard should exist (but as long as people want to re-use existing receptacles...)
 

don_resqcapt19

Moderator
Staff member
Location
Illinois
Occupation
retired electrician
....
While I have not seen a melted 14-50 in person, pictures of melted ones are all over the Internet:
Yes there are, but like everything else installers tend to buy the cheapest available product, and not a higher quality product that would be more expensive.
...

Would there be appetite within the NEC code committees to require a child lock in service cover for this install, especially when installed at child height?
Probably not without a "body count". I have not seen reports of a real world issue here.
To require that "smart" connected devices monitor the temperature of the plug?
Is there even a product listing standard for such a device?
Has there been any NEC discussion of IEC 60309-1 and IEC 60309-2 connectors for this purpose?
I am not aware of any public inputs to require that, and even if there were, I expect the CMP would stick to NEMA standard products.
The public input process is complete for the 2026 code and the work on the first draft is well underway. Any code changes would have to wait until the 2029 code.
 

retirede

Senior Member
Location
Illinois
So I know the NEC does not directly regulate what gets plugged into what outlets, but that said there's a connection.
It seems like NEMA 14-50, borrowed from range and dryer circuits, is becoming the standard for EV chargers. Yet there are some issues:

  1. 1) The available plugs & connectors are rated for a low number of cycles, yet some EV owners travel with their charger.
    2) It’s possible to get electrocuted when plugging in or removing, due to lack of shrouded pins.
    3) The neutral adds 30% or so to copper costs, yet is not needed for EVSE equipment.
    4)The presence of a neutral opens up the chance someone will buy an unfused 14-50 to 5-20 outlet adapter.
    5)When the cord is unplugged, children can access the pins.
    6)There's no reliable make before break for the power pins.
While I have not seen a melted 14-50 in person, pictures of melted ones are all over the Internet:

What do people think of 14-50 as the connector for this purpose long term?

Would there be appetite within the NEC code committees to require a child lock in service cover for this install, especially when installed at child height?
To require that "smart" connected devices monitor the temperature of the plug?
Has there been any NEC discussion of IEC 60309-1 and IEC 60309-2 connectors for this purpose?

A few comments:
1) I don’t believe a significant number of drivers use the same EVSE in their garage and for travel. That doesn’t mean that this isn’t a problem.
2) This is largely mitigated by the requirement for GFCI protection of the receptacle outlet since NEC 2020.
3) True
4) I think you mean 5-15, but your point still stands. It’s even worse if there’s no neutral run to the 14-50R. There are all kinds of unsafe adapters out there. You can’t fix stupid.
5) True of most all plugs used in a dwelling.
6) ???

It’s my opinion that EV charging stresses 14-50Rs in a way that other applications do not. The failures are real. The UL testing standard for temperature rise is inadequate in that only new devices are tested. Long-term effects of thermal and mechanical cycling are not adequately accounted for.

Anecdotally, it appears that higher quality 14-50 receptacles do not experience these issues. Their cost is 5 to 10 times the Leviton, etc. found in big box stores and even supply houses. Since most installations are either DIY or done by the low bidder, chances are the cheap ones get installed.

A child lock requirement doesn’t address the biggest issue - overheated receptacles. I’m not aware of a single incident of child electrocution from an EVSE installation.

Edit - I echo what Don said!
 

suemarkp

Senior Member
Location
Kent, WA
Occupation
Retired Engineer
Most people that mention melting of 14-50's say a 6-50 is better. But your pictures are of melted 6-50's. The GFCI requirement mitigates most of the other points.
 

brycenesbitt

Senior Member
Location
United States
I meant that: come connectors, including J1772, have a pin that breaks or makes first, thus shutting power. Then, the ground pin is longer such that it remains connected until well after the power pins have cleared (including rocking). The NEMA non-locking connectors definitely don't have such features.
--
The GFCI requirement in the 2020 NEC could be seen as mitigating the lousy connector suitability, sure. But that's putting all eggs into a single basket (and creating the much-discussed potential that the dumb CFCI device will result in an empty EV tank in the morning).
--
I think it's only a matter of time before we see a child electrocution due to plug connected EVSEs, given that they are often mounted at child height in areas where children may play.
 

brycenesbitt

Senior Member
Location
United States
Is there even a product listing standard for such a device?
It's not a separate product, it's integral to the load device. For example this image is shows an Aptiv EVSE with extra wires for a thermistor buried in the plug end. A separate listed device would likely to be pre-IoT with no adaptive behavior.

Trust me it sucks when someone drills your gas tank overnight.
It such just as much when IoT-blind shutoff devices prevent your EV from charging. The Aptiv just reduces charging current until the temperature comes down.
 

Attachments

  • PXL_20231124_193115107.jpg
    PXL_20231124_193115107.jpg
    319.6 KB · Views: 17

brycenesbitt

Senior Member
Location
United States
A 14-50 plug melt, just for reference :)
A 5 cent thermistor in the plug, sensed by the IoT EVSE, would put an end to photos like this.

People could still drop metal objects into the plug pins, however, if the plug is out from the wall.
The NEMA plugs are not protected against that at all.
 

Attachments

  • DA7EDF7F-FF1C-46C1-BF10-858289F398DB.jpeg
    DA7EDF7F-FF1C-46C1-BF10-858289F398DB.jpeg
    298.1 KB · Views: 8

retirede

Senior Member
Location
Illinois
A 14-50 plug melt, just for reference :)
A 5 cent thermistor in the plug, sensed by the IoT EVSE, would put an end to photos like this.

People could still drop metal objects into the plug pins, however, if the plug is out from the wall.
The NEMA plugs are not protected against that at all.

When only one of the two current-carrying connections is damaged (like this picture) I immediately suspect that pole of the receptacle had its wiring termination improperly torqued.
That’s not a given, of course, but a good possibility.
 

retirede

Senior Member
Location
Illinois
It's not a separate product, it's integral to the load device. For example this image is shows an Aptiv EVSE with extra wires for a thermistor buried in the plug end. A separate listed device would likely to be pre-IoT with no adaptive behavior.

Trust me it sucks when someone drills your gas tank overnight.
It such just as much when IoT-blind shutoff devices prevent your EV from charging. The Aptiv just reduces charging current until the temperature comes down.

Other makes also incorporate this feature (I know Webasto does). I could see making this part of the listing requirement for cord-and-plug connected EVSEs.
 

retirede

Senior Member
Location
Illinois
Example plug connected fire: and note how well the plastic box held up!
https://www.reddit.com/r/evcharging/comments/17zeha4

Interesting facts given - 30A EVSE, 40A OCP, 6-50R. Theoretically, there’s no way that should have happened if everything was installed properly. 30A should not be able to overheat a 50A receptacle.

Observations - one pin seems to have experienced a good deal more heat than the other….undertorqued?
Remnants of a wirenut in the box - possible failure point?
I think the back of the box is gone - isn’t that yellow showing building insulation?
 

don_resqcapt19

Moderator
Staff member
Location
Illinois
Occupation
retired electrician
Interesting facts given - 30A EVSE, 40A OCP, 6-50R. Theoretically, there’s no way that should have happened if everything was installed properly. 30A should not be able to overheat a 50A receptacle.

Observations - one pin seems to have experienced a good deal more heat than the other….undertorqued?
Remnants of a wirenut in the box - possible failure point?
I think the back of the box is gone - isn’t that yellow showing building insulation?
Why not? It does not take much resistance at the connection points to start creating a lot of heat when the load is drawing 30 amps. I²R

In many cases, when I had to work on damaged receptacles, I found that the real fault was the internal connection between the flexible cord and the plug blade that transmitted the heat into the receptacle. Not saying that is what is going on with the EV cord and receptacle connections, but just something I have experienced on other high current cord and plug connections.
 

retirede

Senior Member
Location
Illinois
Why not? It does not take much resistance at the connection points to start creating a lot of heat when the load is drawing 30 amps. I²R

In many cases, when I had to work on damaged receptacles, I found that the real fault was the internal connection between the flexible cord and the plug blade that transmitted the heat into the receptacle. Not saying that is what is going on with the EV cord and receptacle connections, but just something I have experienced on other high current cord and plug connections.

My line of thought is that if the plug/receptacle combination is designed and listed as a 50A device, a 30A current flow should never overheat it unless something else is amiss. The bad connection in the plug you describe could be that something.
 

Carl_in_AZ

Member
Location
Disney9405#!
Occupation
Retired Electrical Engineer from SQD, T&B, Leviton and Cummins
From my days working for Square D, T&B, and Leviton as an Electrical Engineer I have seen what cyclic loads created by inverters (chargers) do to a connection point. If the camber on the dryer receptacle is not beefed up or the breaker in the load center is next to another heavily loaded breaker the receptacle's blades/pins will eventually start to pit and create a fault. Also, heat in breakers feeding cyclic will cause the thermal units in the dryer/charger breaker to trip. This is typically not an issue in panelboards where the breakers have space between them. Some cities are looking at changing requirements for breaker locations in household load centers when feeding chargers. I plan to use the Hubeell or Bryant dryer receptacle after seeing other issues in the Phoenix Valley. Keep in mind electric dryers only see resistance loads, not cyclic loads so all receptacles work well with electric dryers.
 
From my days working for Square D, T&B, and Leviton as an Electrical Engineer I have seen what cyclic loads created by inverters (chargers) do to a connection point. If the camber on the dryer receptacle is not beefed up or the breaker in the load center is next to another heavily loaded breaker the receptacle's blades/pins will eventually start to pit and create a fault. Also, heat in breakers feeding cyclic will cause the thermal units in the dryer/charger breaker to trip. This is typically not an issue in panelboards where the breakers have space between them. Some cities are looking at changing requirements for breaker locations in household load centers when feeding chargers. I plan to use the Hubeell or Bryant dryer receptacle after seeing other issues in the Phoenix Valley. Keep in mind electric dryers only see resistance loads, not cyclic loads so all receptacles work well with electric dryers.
Does anyone have a graph of what the current
for a typical car charger looks like ?
 

wwhitney

Senior Member
Location
Berkeley, CA
Occupation
Retired
Does anyone have a graph of what the current
for a typical car charger looks like ?
A car's on-board charger will generally have a maximum current of 27.5A, 30A, 32A, 40A, or 48A (other values are possible). A 208V or 240V EVSE will generally have a maximum current of 16A, 24A, 30A, 32A, 40A, 48A, or rarely larger. The EV will draw min(charger rating, EVSE rating) for many hours, and only start tapering below that for the last hour.

So a long plateau for 0-20 hours (depending on EV battery size, state of charge, and charger rate) followed by a gradually decreasing tail for maybe an hour.

Cheers, Wayne
 

Carl_in_AZ

Member
Location
Disney9405#!
Occupation
Retired Electrical Engineer from SQD, T&B, Leviton and Cummins
A graph does not show the cyclic current. You will need a scope read out. I have included a link to an example I found on the web which is similar to what I have seen in the past.. I am working as a volunteer with the DOE on electrical charging infrastructure. I will see if they have an up-to-date read out.
 
A car's on-board charger will generally have a maximum current of 27.5A, 30A, 32A, 40A, or 48A (other values are possible). A 208V or 240V EVSE will generally have a maximum current of 16A, 24A, 30A, 32A, 40A, 48A, or rarely larger. The EV will draw min(charger rating, EVSE rating) for many hours, and only start tapering below that for the last hour.

So a long plateau for 0-20 hours (depending on EV battery size, state of charge, and charger rate) followed by a gradually decreasing tail for maybe an hour.

Cheers, Wayne
I was referring more to what carl said in post #19. These are certainly switching power supplies and not a simple rectifier that "chops off" half the wave so their Waveform will be complex.
 
Top