Why fast charging slows down
Every charging curve on this site has the same shape. A rise, a plateau, then a long fall. The fall is not a defect and it is not the charger being polite. Three separate mechanisms push it down, and they take turns.
1. Lithium plating risk, which rises with state of charge
Charging means moving lithium ions from the cathode into the anode’s graphite layers. An empty anode has plenty of space and accepts them quickly. A nearly full anode does not.
Push current into a full anode and the ions cannot insert fast enough, so metallic lithium deposits on the surface instead. That metal is gone from the chemistry forever, which is capacity loss, and it grows in needle shapes that can eventually pierce the separator and short the cell.
So the battery management system walks the current down as state of charge rises. The taper above roughly 60 to 80% is an insurance policy against permanent damage, and it is the reason the last 20% of a charge costs more time than the first 70%. In the calculator, a Model 3 goes from 10 to 80% in 24 minutes and from 80 to 100% in 28.
This is also why charging to 80% is the road-trip convention. It is not superstition, it is where the curve stops paying.
2. Heat, which the coolant loop has to remove in real time
Resistive loss in a pack goes as current squared. A cell absorbing high current makes heat faster than a cell absorbing low current, by a lot, and the cooling system has a fixed capacity. When the cells approach their thermal limit, the BMS derates.
Cold is worse, and less intuitive. In a cold pack, ion mobility collapses and internal resistance rises, so the safe charging current drops dramatically. A pack at freezing can accept a third of its rated power or less until it warms. Modern cars heat the pack on the way to a charger when the navigation knows where you are going, which is why “preconditioning” matters so much: it is the difference between arriving ready and spending the first fifteen minutes warming up on the charger’s dime.
Try the cold toggle in the calculator. A cold Model 3 needs about six extra minutes for the same 10 to 80%.
3. The cabinet’s ceilings, which have nothing to do with the battery
The third reason is not chemistry at all. Power is volts times amps, and the cabinet has a limit on each.
Most 350 kW CCS cabinets stop at 500 A, so a 400 V-class car is capped near 175 to 200 kW regardless of the sticker. An 800 V car draws the same power at half the current and gets the full rating. Tesla’s cabinets went the other way, roughly 700 A at only 500 V, which suits a 400 V Tesla and throttles an 800 V car to split-pack charging.
These are the details covered in the 500 A problem and Supercharger reality.
They take turns, and the calculator says which
In any given session one of these is the binding constraint, and it usually changes partway through. The calculator names whichever held for most of the session: the car’s taper curve, the charger’s current limit, its voltage ceiling, its power rating, a cold battery, or the station’s own battery if there is one behind the dispenser.
That single line is the most useful output on the page, because it tells you what to change. A cabinet-limited session gets better with a different cabinet. A taper-limited session does not get better with anything, and you should unplug at 80% and drive.
Next
The 500 A problem
Why a 400 V car cannot use a 350 kW charger, and why the 250, 260, 320 and 350 kW stalls all give it exactly the same session.