Why did my EV charge outlet overheat and burn the wires?

Question

I bought an EV that came with a free level 2 charger. I had a Honda “certified electrician” come out and install a Honda re-badged JuiceBox on an unused 60amp two-pole breaker (previously used for a hot tub).

Everything worked well for roughly 11 months, until last night. I came out this morning to find my car at less than a full charge and the breaker tripped. I stupidly thought I could just reset the thing and be on my way, but when I flipped it, the whole box sparked up. The attached pictures shows the aftermath.

Anyone know what went wrong with this? Is it safe to say this was an installation issue, or am I off base? I’m assuming my EV charger is fried as well.

Answer 1

Series arc fault on the extension

Is the short answer. There was a poor connection (where is obvious: it's Ground Zero of destruction) and that connection started to arc. Similar to arc welding, arcing makes a LOT of heat.

Looks to me like a wire nut inadeqately torqued. They require "gorilla tight, not monkey tight", especially the big ones. One problem with wire nuts is the inability to specify a screw torque - which a) science has revealed is important at all terminal sizes, and b) electricians are proven to be bad at.

I see absolutely no twisting action on that nut. People often believe wire nuts are "caps" only there to insulate. The twisting is essential to crush the wires into each other.

Contributing factor #1: the splice

Wire problems almost always happen at terminations. When a hardwire connection is continuous from breaker to EVSE ("charger"), it has thermal protection on both ends. Most EVSEs have a thermosensor at the power terminals and are actively monitoring temperature there. This arrests arcing before it can be serious.

Breakers have a passive way of monitoring terminal temperature. They have 2 modes of trip: an instant-trip operated magnetically, and an inverse-time delayed trip operated by passing current through a bi-metal strip. Thermal rise in a wire is the square of current, so this works nicely. When the terminal screw on a breaker heats up, that heat is telegraphed along the current pathway to the bimetal strip, heating it externally and causing the trip.

This usually happens only after enough arcing has happened to make the problem very obvious.

However with a midpoint splice, there is no thermal management at that splice.

Contributing factor #2: The Fastest Charge Possible (tm)

Technology Connections covers this beautifully in their "Guide to home EV charging" video (a must-watch)...

"Don't just go for the Fastest Charge Possible: It's expensive, and may come with headaches you didn't bargain for." And in the experience of the EV community, that's usually this. Such extreme power, especially when run at thermal limits of the wire, WILL find every flaw in the wiring and make it crispy.

Let's do some quick math on home charging. If you go 15,000 miles a year, doing a mediocre 3 miles/kWH, that's 5000 kWH a year that needs to go into the car. That is 13.7 kWH per night. Assuming conservatively 10 hours a night of charging with no downtime on weekends to catch up, that's 1.37 kWH/hour == 1370 watts. Regular old "level 1" charging is 1440. So that works out on the average especially with a little more than 10 hours a night. The only reason most bother with "level 2" charging is imagining transient events bigger than battery capacity.

So one way to mitigate this is simply turn down charging amps. Normal wire heating is the square of amps, so e.g. a drop from 48A to 24A (still overkill, still fills the car overnight) will reduce heating by 75% and reduce the degree to which arcing starts.

Indeed we've seen multiple reports of people who installed charging for their plug-in hybrid (which pulled 12-16 amps) and it served them well for years, until they brought in a battery EV which pulled 32A and it caught fire the first night.

Not OP's problem, but sockets exacerbate this problem - that's why we warn against them generally and cheap under-$30 range outlets in particular.

Contributing factor #3: No arc-fault detection

Arc fault detection involves analyzing the AC voltage and current waveform to look for telltales of arcing - think of that "crinkle-crunch" sound of hooking up speakers with the amplifier turned on. The algorithm suits GPU style microprocessors, such as those in video cards, Bitcoin miners and AI.

Development of affordable AFCIs didn't really line up with the timeline of the EV charging standards being developed. In addition, the industry was (and is!) being pressured to reduce the parasitic power usage of EV chargers e.g. Energy Star. So there's no market interest in providing series arc fault detection in the "charger" (EVSE). Thanks, Energy Star!

You can do it at the circuit breaker, but that segues us back to the "how fast do you really need to charge" discussion, because the largest generally available AFCIs are 20 amps, giving 3.84 kW charging rate, half the 7.7 kW rate of typical mobile chargers and 1/3 the maximum of wall units. And that brings out people's anxieties.

Contributing factor #4: Box not grounded.

Code requires a metal box be tied to ground. I see only 2 ground wires in that nut, so clearly the (hack?) electrician didn't run a pigtail to the box. I would've mounted actual ground lugs on the grounding screw.

Had they done so, the first wire to be bared by insulation melting would have contacted the box and tripped the breaker a lot sooner. As things were, it energized the box until melting proceeded far enough to melt the OTHER hot wire and short them.

Answer 2

It's hard to tell from the photo but (assuming 90c rated wire) this installation requires #6 awg copper (or larger) for the phase and #8 for the ground (which looks correct).

If the wire is too small, that's your problem. If the wires are properly sized #6 copper then it looks like they were improperly joined in the wire nut.

That nut looks like an Ideal brand "big blue" wire nut. Edit: I was mistaken. Assuming it's an Ideal big blue, it should be fine for 2x #6 wires but the wires in your box, specifically, still do not appear to have been properly bonded in the nut.

The joint in the nut failed, leading to arcing and/or heating. The heat melted the wire, which then shorted to ground in the box, tripping the breaker.

Run new conductors and join them properly. This is not required but I suggest a larger box, too.

Edit: additionally the junction box doesn't appear to be properly grounded. The grounding conductor must be bonded to the metal junction box.

Answer 3

I would blame installation, or incorrect provisioning of your charger that allowed it to draw more than the rated full-duty rating for a 60A circuit (which is probably around 48A or so for full-duty appliances/chargers).

It's obvious to me from the pictures that at least one of the wire nuts did not do a sufficient job of connecting the wires in the junction box and had a relatively high-resistance connection which allowed extreme heating to occur with the amount of current being drawn by your charger. I would guess it's the one completely burned in the lower right, as the other nuts (from what I can tell from your pictures) look to be in somewhat better shape.

The extreme heating could have been from over-provisioning your charger allowing to draw too much power, or poor connections.

I can't tell whether your wire is copper or aluminum (at least some looks to be aluminum, but that could just be extreme heating of the copper). With high amperage wiring, use of aluminum wiring is fine, as long as it's properly joined with aluminum-rated connectors and techniques. It's possible that the electrician didn't know (or care) to use proper techniques for joining aluminum. Improper techniques would then allow connections to heat up excessively, causing damage like you have in your situation. But it's impossible to know from a distance what the exact cause was in your case.

The resolution is to get an electrician out to repair the damage and make sure the provisioning of your charger doesn't allow it to draw more than the full-duty rating of the circuit. There's a good chance your charger will be fine, once the wiring has been repaired.

Answer 4

Its hard to see from your image. It looks like the box experienced an arc to me. (PS this is why you only splice wires in a box, the box did its job).

But there are 2 things wrong that I see.

1. The wire nuts aren't tight enough. it looks like the wire nut for your red wire wasn't tight enough. Your ground wires (the green ones) don't appear to have been tightened particularly well either. I can't see the wire nuts for any of the other wires, but I assume that they are loose as well.

You can tell that they aren't very tight because the wires didn't twist around each other at the entry point to the nut. This is telling sign that the wires aren't tight. I believe that the "certified electrician" tightened them by hand isntead of using pliers. Hand tight isn't very tight. EV's pull a lot of current which causes heat, the heat causes the wires to expand which with a loose nut puts just enough distance between the wires as after they cool and shrink that you they can create an arc.

2. Your box isn't grounded. The ground wires should have a pig tail connecting to a grounding screw in your box.

How do I fix this?

Use a WAGO next time, even though they cost more than a wire nut. Its like 50 cents vs 5 cents. This is notably less than the cost of a home fire. Why use them? It is much easier to install a WAGO correctly then it is to install a wire nut correctly.

If you can't tell, I tried using a WAGO once, and fell in love. I personally think they end up being cheaper to install because it takes less time to put in a WAGO then it does to use a wire nut, and time is money. I just learned that WAGOs don't come in 6 awg.

Use a polaris style connector you will want to use anti-ox with it.

Both of which are easier to tighten correctly than a wire nut.

Answer 5

There’s a lot of bare copper there; maybe some contact between hot and neutral/ground, maybe via the ungrounded box itself? Sounds like your ground did the right thing the first time around (tripping your breaker), but the second time around, when you flipped it back, another contact was made that toasted the box. Could be install issue though if the charger caused the breaker to trip, and that did something in the box, who knows..

Answer 6

A short circuit (high current) made the breaker (aka MCB) break. When the short circuit current is high and sudden, then the breaker can get partly or fully damaged preventing it from breaking as expected in case of a new incident. A MCB has a thermal sensor, which is slow to break, and a solenoid which is very fast. Even if the breaker did break again after you switched it in, then the protection could be too slow leaving your charger circuit partly unprotected for a while (seconds).

Now, what happens if you have an unprotected short circuit? The weakest point will act like a fuse, and/or your electric wires will get hot and the insulation will melt or even burn. Your photos shows a clear evidence for the later hypotheses. It seems like the fire started in the right part of the wires. (Perhaps closest to the charger?) Perhaps the wires are fixed or pressed close together at the outlet in the right side of the box?

When the insulation melts, then the copper wire touches each other first at the point having pressure, and you get a new shortening here. Closer to the breaker I guess, and evidently with a much lower resistance than the first short circuit. A full (copper against copper) unprotected short circuit is almost always observed like an explosion especially inside a box. (Remember to care for toxic gases)

The blue plastic parts on the photos look like wire assemblies. In many countries regulations demand unbroken wires without connectors/splitters between the switchboard and the EV charger. Connections have higher resistance than unbroken copper and become heated faster. A fire is expected to start here.

But where did the first short circuit take place? If you charger actually is a charger and not just a charging cable then it should be fully protected against common errors by internal fuses/breakers. That includes errors in your car and the cable from your charger to the car. An error could be caused by over-voltage in the net (did you observe thunder during the night?) I do not expect any serious damages to the charger.

Anyway, I suggest you let the manufacturer check it. They have an obligation to do that because they participate in the UL system.

A more likely explanation is damage to cable insulation caused by rodents. All outdoor cables must be protected by metal shields if they are exposed to attacks from rodents. That typically means if they are close to the ground.

What you need to do now:

• The breaker should be locked (preventing someone from switching it on), and you should have it replaced soon.
• All wires in the pipes on the photo should be replaced because the insulation could be damaged by the heat.
• IEC 60364-7-722 (EV charging) usually requires that you have a surge protective device (SPD) in your switchboard. Let the electrician check that.

What did you learn? Always check your cables/installations/equipment before you switch a released breaker back on.