[bryanlarsen]

The demand for energy isn't changing. If you're charging 5 times as much, but only charging 1/5th as often, you aren't taking more from the grid, you're just making demand on the grid more inconsistent.

There are two main ways of handling that randomness.

1. Spread it over a large number of vehicles, and let statistics average out the demand spikes. We're doing this at a significant rate, adding > 1 million EV's to American roads every year. Also, we already have superchargers with 98 stalls. There's going to be little demand difference between 98 350kW chargers vs 32 1MW chargers. 9 350kW vs 3 1MW chargers loads a grid harder, but 98 vs 32 not so much.

2. Batteries. Charging stations are already large enough that many are utilizing batteries to smooth electricity demands.

[7e]

Superchargers with 98 stalls, today, absolutely cannot handle peak charging of all those stalls at the same time. That’s the equivalent of 34,000 homes turning on at the same time, which is basically a medium size city.

[bryanlarsen]

Cannot handle it, and doesn't have to handle it. A Tesla will only pull the full 350kW for 2 minutes in a charging session, and sessions average ~20 minutes. And even when the station is full, there are always stations pulling 0W because they're switching between vehicles et cetera.

But if it did need to handle that level of instantaneous load, it would be straight forward to supply. 98 Tesla's worth of batteries on site could do it. The grid tie is substantially smaller, because the grid tie only has to handle average load, not instantaneous load.