Which is, as I said, NOT determined by any single car. You will almost always have the thing running at full tilt for 15 minutes at least once during any given month. Yes, in theory, you might get really lucky once in a while, but in practice, the demand charge for a supercharger in any given month is almost guaranteed to be based on its maximum utilization, +/- a few percent.
But the next thousand Model 3s during that month add zero additional demand charges. And if that first car charges at a slower rate, the second one will add those same demand charges, which means it doesn't save Tesla money if that one car charges at a slower rate. Demand charges only matter if there are NO cars charging at a higher rate. In other words, it depends almost entirely on the speed of the charger, not on whether some cars come in mostly full or not.
And even if the cost of leasing parking spaces and building the charging equipment were zero, Tesla would still not be able to build enough chargers for everyone to use only the top 20% of their batteries, nor would they be able to convince users to spend that much extra time supercharging even if they could. The whole advantage of the supercharger comes from being able to charge quickly. So there's no feasible universe in which they won't always be paying approximately the maximum demand charge every month in any area where the superchargers get reasonable amounts of use (and in areas where they don't, they also probably are not paying demand charges).
First of all, it isn't 200 kW per stall. Even at a V3 supercharger, you only have about 125 kW per stall, which is only half again more than a V2 supercharger. The power is just shared across a lot more stalls, so you are much less likely to be limited because of the car next to you. An 8-stall V3 supercharger has a 1000 kVA transformer, so at $20 per kW, that's only half as much peak power as you seem to be expecting.
Second, your demand estimates are way too high, at least when we're talking about industrial quantities of power. Let's take PG&E for a quick example, because it is where several of the most expensive superchargers in the U.S. are located. With a 1000 kVA transformer, an 8-stall supercharger is almost guaranteed to be on PG&E's E-20 service, with a primary feed. So their demand charge is only $5.55 per kW — about a fourth of your estimate — at least as of the Jan. 1., 2020 rate plan.
So with real numbers, that comes to only about $5,550 in demand charges.
And those superchargers often see 80% occupancy from 7 A.M. to 10 P.M. If we assume an average of 60 kW per car over the charging interval (best guess) at 80% occupancy, even if you assume that they are completely idle the rest of the time, that translates to about 5,760 kWh of power consumption per day, or 172,800 kWh in a 30-day month. If we divide $5,550 by 172,800, that means they're probably paying only about 3.2 cents per kWh because of demand charges.
Demand charges are a big deal for small charging stations with low utilization. For superchargers, they really don't matter, because the utilization is so high.
The 50 kW Tesla car owner doesn't exist. What you have are cars that, entirely by fluke of being at a particular state of charge, happen to draw less power. Tesla cars (ignoring throttling) support a peak charge rate of 120 kW on the 60D, and even higher on all the others, AFAIK.
And the reason I want to charge the slower-charging car more is because those demand charges won't go down merely because one particular car charges more slowly, but the amount that Tesla spends on maintaining that charger while that car charges more slowly are higher for that car than for a car charging at a faster rate.
That said, I'd be perfectly fine with Tesla charging an extra 1.5 cents per kWh in demand charges for using the V3 superchargers. I'm not sure anybody will care about that relatively tiny cost difference, and it probably isn't worth the extra hassle to make their billing system handle it, but I'd be fine with them doing that.
Last edited:
