The SOC that the car displays, also called SOC from the CAN, is using the exact formula you showed in your earlier post. Assuming that the display and CAN SOC were 54%, then the SOC that you would use for the remainingkWhnom value would be about 56% ( The same calculation with the buffer from the numerator and denominator removed). I think this is called SOC UI from the CAN. In that case, the remainingkWhnom would be around 41.3 kWh, not 40 kWh, assuming a 3.3 kWh buffer like you were using in your numbers.
So I don’t believe you need a new charge constant to explain the CAN data, but it is difficult to understand exactly why Tesla does certain things the way they do.
You're right. It's true that formula may work always for fullkWhNom; I am just saying it can't be extrapolated to OTHER states of charge; you can't conclude that remainkWhnom = remaining rated miles * charge constant.
That formula definitely just does not hold - pretty clear from the CAN bus readbacks posted here that each rated mile is
not equal to the charge constant Wh value (245Wh/rmi). That was really my only point I guess. And you're right, it may not be coincidence at all. That may well be the exact formula.
A contradicting point though: some data posted around here has indicated that a new full battery gives fullKWhNom = 77.9kWh (or something very close to that). There's no way to explain that - for an LR AWD that would be 318 rated miles at 100% - and I've never seen that. Nor has it been reported that such a battery stays "stuck" at 310 miles for a period of time of driving (so it's not a display cap). So that datapoint would not match that formula.
Worth stating again though: For all these numbers, just because the formula, the trip meter, the CAN bus, etc., say that the capacity of the battery is 76kWh or 78kWh (or whatever), it does not mean that's the actual battery available capacity - obviously the actual capacity could be something entirely different - and very likely it is different, as it will of course
actually match the EPA 79kWh after 4000 miles (with some car-to-car variation). It's all just a harmless scaling factor of course.
Regarding the SOC, based on what I see above, what the CAN bus reports for SoC (even the value of SOCui) never matches what the car UI displays. I don't have a CAN reader so really have no idea; just basing on reports above. But there is no CAN bus "SoC" reading which matches what the car shows - regardless of which SoC you're talking about. (There were four values from 55.8% to 60%; actual UI reading was 54% (close to 55%).)
Regarding your formula suggested - yes - not sure why I did not conclude yesterday that SOCmin is closest to the formula remainkWhnom/fullkWhnom...but that certainly is a possibility for the formula. It's not perfect though. What the other various SoCs reported are, I have no idea. But none of them match the car UI.
The SoC in the car appears to always simply be the formula I gave above (remainingkWhNom - bufferkWh) / (fullkWh - bufferkWn), of course.
Anyway, picture becoming a bit more clear...just a few more readbacks at different states of charge with adequate documentation of the exact state during the readback, and some comparison to the trip meter, and it will all become clear about how the trip meter, CAN bus, and charging screen relate.