Another way (from first principles): Assume you buy two cars per driver. The driver parks one car at their solar panels at a time, so one is available for use 100% of the time, and at least one can charge off the panels 100% of the time.

Assume a 10kw solar system with no batteries, but with a level 2 charger. That costs $28K this year. Assume $50K per car. The system cost is $128K.

Assume the climate is such that you can charge the cars at 6kw (max output of the charger) for an average of 8 hours a day (pessimistic in summer, optimistic in winter).

This setup should last about 10 years. (Or sell the cars after 5 and get money back for new cars.)

That’s 365 * 10 * 8h * 6kw usable for the cars, or 175.2MWh, giving us $0.73 per kWh. Clearly the sky is falling. I’m going to get a steam engine for my buggy!

I forgot to figure the depreciation of the cars. We wasted one because this scheme is dumb, so the depreciation for an equivalent ICE car would be zero. For the other car, I think it’s reasonable to assume 90% depreciation. Say the ICE car depreciates $40K. We can sell the two EVs for a total of $10K. Now the total cost (sans car) for the system is $78K, or $0.445 per kwh — cheaper than California’s grid.

I forgot to figure interest on the $78K of capital. At 8% average return, that’s a bit over 2x, getting it closer to a dollar per kwh.

For a 4 mile/kwh car, that’s $0.25/mile. Of course, if you assume the existence of civilization, then the price drops a lot. For instance you could only buy one car, and you could size the solar smaller, or also plug the house into it.

Anyway, in my hypothetical mad maxian hellscape that’s experiencing healthy, steady economic growth and has access to cheap refined gasoline, he’s still off by a factor of 2.5x.