@SageBrush is right, people are not consistent in how they report this since there are so many ways to do so.
Sorry, I'm not super succinct. But hopefully this is more information than you'll ever need!
Using changes in percent (state of charge) is one of the worst ways to compare, but happens. It's bad because a percent contains different amount of energy (especially LR vs. SR+), but also because of degradation as you mentioned.
With my low-degradation LR pack at the time, leaving from a warm-ish garage and thus battery (usually 10-12C) these were my observations:
- Around 10C, 50F: Derate 20% compared to rated range
- Around 0C, 32F: Derate 30%
- Around -10C, 14F or colder: Derate 40%
You
cannot take those numbers straight across to an SR+! As the efficiency loss is primarily due to heating, there is no substantial difference in the cabin between the SR+ and the LR. Therefore, the same
energy is required for heat the cabin given some set of conditions (how cold, speed, etc.). And that same
energy requires more
percentage from the SR+ pack than the LR, because the SR+ has less energy.
Further, note that derating
range by 30% is not the same as 30% higher consumption. It's the inverse: so 142% relative consumption, and thus 42% "more" consumption. Doubling consumption would half the range, which shows the inverse relationship.
For example, at freezing with the 30% derate on LR, I can figure...
- My car is normally rated for 234Wh/mi, but would consume 334Wh/mi at 0C, making 100Wh/mi roughly required to heat the cabin while driving. (I promise I didn't arrange for that to be such a round number, weird)
- The 2020 SR+ is rated for about 201Wh/mi. Adding 100Wh/mi gives 301Wh/mi. This is effectively a 33.2% derate for the SR+ in the same conditions, where the LR is only 30%.
- Note: This comparison looks better than it otherwise would, because the LR is already rated for worse efficiency than the SR+. I was actually surprised by the calculation, and assumed the derate would be higher.
What I want to highlight is that while the derate number is worse, the same energy is being used to heat the cabin in both cases. One is not less efficient at heating than the other. You just
have less energy in the SR+.
This is why early or small-capacity EVs can't go anywhere in winter. The cabin heating requirements are roughly the same, but their batteries are
much smaller. Cabin heating becomes their primary power draw over their small locomotive needs. A 100km-rated little EV work truck might only go 20km in the winter (anecdote from a family friend).
Now, like I said, percent is a bad way to do this. Using those efficiency numbers is better, because they can be adapted to differently-degraded batteries. Problem is you now need to know 1.) the efficiency constants for your car, 2.) how much energy you have usable, and 3.) adapt those percentages above to Wh/mi or Wh/km. All entirely doable, but they're quite a bit of work for something that's plus or minus 10% anyways.
Actually, let's do that.
- 10C, 50F:
- Additional 59Wh/mi, or 36Wh/km
- At 80km/h or 50mph, implies 2.9kW of heating
- 0C, 32F
- Additional 100Wh/mi, or 62Wh/km
- At 80km/h or 50mph, implies 5.0kW of heating
- -10, 14F, or colder
- Additional 156Wh/mi, or 97Wh/km
- At 80km/h or 50mph, implies 7.8kW of heating
- This number is not realistic for just the heater. Heater is understood to draw 7kW max. This may include AC (for defrosting) or simply have additional factors that start to become apparent. Happily, this means it's sort of the upper bound on heating impact, which would explain why it's the worst number I observed despite being in much colder temperatures.
Those numbers might sound like a lot for steady state, but the Model 3 is not well insulated. You can feel the cold off the glass even when the car is otherwise warm.
Note: I haven't found that the "derate" numbers above change much with regard to speed. Heating power is required to just be present in the cold. The faster you get somewhere, the less total energy you use for heat (since you heat for less time), but the more you used for driving, and vice versa. This seems to somewhat cancel out, though I haven't done any super scientific testing on this. That said, if you're going incredibly slow on average (less than 10mph, let's say), then it might be possible to get worse numbers.