Hi.
I just leased a 2016 eGolf SEL yesterday.
One of my claims to "fame" is that I design and build EVSEs as a hobby. One of my creations is called the
J1772 Hydra. Until Chargepoint came along, it was the only double-headed, single-circuit EVSE in existence. It's still the only one that can be used to safely "split" a J1772 charging station to charge two cars at once.
All that is merely my way of saying that I
have read the J1772 specification and am intimately familiar with it.
The specification is oriented more towards readers that are intending to implement an EVSE, rather than those intending to build a car.
I can't speculate on VW's implementation without further analysis of how the eGolf behaves, and I'm not 100% sure I am interested in going down that road, but for the purposes of some of the discussion above, there are two signals that are the most relevant for discussion.
The first is called "Proximity." In general, this signal is not routed to the EVSE (the Splitter variant of the Hydra uses it, however, to pass inlet proximity transitions to the vehicle, but that's the exception to the rule). It's a resistor divider signal that's connected to a switch in the latch release button of the J1772 plug. Pressing the button to unlatch the physical latch that holds the plug in place is supposed to cause the resistance on the proximity pin to change. The spec allows the vehicle 100 milliseconds to stop charging and reduce current draw to "low" levels (a few hundred mA) when a proximity transition occurs.
The second signal is called the "pilot." It is a bidirectional communication channel between the EVSE and the vehicle. The EVSE sends a +/- 12 volt 1 kHz square wave to the pilot through a 1 kilo-ohm impedance. The duty cycle of this square wave is an indication to the vehicle of how much current it is allowed to draw (this is independent of the voltage of the EVSE - it is merely a current limit. A 30A pilot can mean 3.6 kW at 120V or 6.6 kW at 208V or 7.2 kW at 240V). The vehicle communicates its presence and status to the EVSE by connecting different resistances between the pilot and ground (with a diode in-between - more about that later). 2.7 kilo-ohms indicates the vehicle is present, 882 ohms indicates the vehicle desires power and 240 ohms indicates the vehicle desires power and requires ventilation (this is largely obsolete - it's for vehicles with lead-acid batteries that can produce hydrogen gas during charging).
The spec allows a vehicle up to 5 seconds to notice that the pilot duty cycle has changed or that the pilot has been dropped entirely.
Typically if a vehicle is actively charging and someone removes the plug, what happens is that they press the latch release button and a proximity event tells the vehicle to stop drawing current. The vehicle does so very probably before the plug even begins to move. Some vehicles will transition the pilot resistance from state B to state A and the EVSE will shut off the power, but this is not required by the spec. If it doesn't happen, then as the plug is withdrawn, the pilot pin - which is shorter - will break contact first, causing the EVSE to shut off. The power draw from the vehicle will already have been reduced to negligible levels by the proximity transition, which will prevent arcing, both in the plug and in the contactor of the EVSE. The power pins will break contact next, followed by the ground pin - the longest one in the plug.
It's conceivable that VW is locking the plug in to prevent proximity transitions because their charger can't meet the spec's mandate to discontinue high current draw within 100 ms. It's also conceivable that locking the plug in is an artifact of CCS charging, which is an entirely different process (one of which I know very little by comparison), although the plug lock still operates the same way on the SE eGolf, which lacks CCS.
I've attempted (fruitlessly) to remove a Blink plug from my eGolf. When I did so, I could clearly hear the proximity switch click in the button, however so far as I could tell, there was no reaction from the vehicle, indicating that the proximity transition was ignored. If that's the case, then attempting to defeat the latch may result in arcing when the plug is removed. I've witnessed a 30A arc when an EVSE is disconnected under load, and it's a... more impressive thing... than you might expect if you've seen, for example, a spark from unplugging a toaster that wasn't off.
If VW is working around an inability to react to proximity transitions within the time allowed by the spec by locking the plug in place, then to my knowledge they're the only automobile manufacturer that has resorted to such. Highly unexpected from such a deeply experienced batch of German engineers.