Thanks for the interesting and thoughtful post! It is substantial food for thought.
. However, the drawback to these resistors is that they are the primary cause of vampire drain.
I kind of doubt this. What is the value of the bleed resistor? Once we know that we can calculate, though...
I guess it is possible that vampire drain (*after rebalancing is complete*) is actually very low on Model 3, but it certainly is hard to determine. My best estimate is around 40W average in an optimized state, but that is derived after a brief period of sitting, and perhaps after longer periods of sitting the average is lower (again, difficult experiment to perform repeatably).
The other comment I would have here is that once the balancing is complete, they don’t bleed anymore regardless of the resistor value, do they? Once the battery is balanced (all bricks at same voltage), there isn’t a need to bleed. So I could see a smaller resistor value leading to greater short term vampire loss (quicker rebalancing), but as far as longer term vampire loss is concerned I can’t see the resistor value as mattering.
On 1/20/2020 at 30,700 miles, I was down to 270 miles full range, which is 40.8 miles lost (15.1 %). The first good, accurate recalibration occurred 4/16/2020 at 35,600 miles and brought the full range up to 286 miles. Then another one occurred on 8/23/2020 at 41,400 miles and brought the range up to 290 miles, now only a 20 mile loss (6.9 %).
Originally you had 78-79kWh.
This is actually 290*245 = 71kWh
And
270*245 = 66.1kWh
So you recovered from ~15% loss to ~9% loss.
This is still substantial recovery, of course.
1. This issue is primarily an indication/estimation problem, not real battery capacity loss.
How would you explain calibrated charging interval experiments showing shorter charge intervals on cars showing loss of capacity? In general, you can estimate charge times within a few minutes. This indicates that the estimates of capacity are close to accurate.
Every time we have really dug into it with calibrated charging experiments here (we know what the charging overhead is from the EPA documents and also from other methods), the car’s estimate of energy available appears to be within 1-2% of accurate.
Still, interesting. I’ll try leaving my car sitting at a low charge for a few hours a couple times over the next week. I really need to measure the discharge energy on the car meter as well from a high SoC - haven’t done that in a while, and now that capacity is degraded it should show a lower value for how much was used for a full discharge. (As we well know, the discharge energy does not capture all the energy used (off by perhaps 1%) - I can’t compare that value to the charge event or the BMS estimate. I’m just talking about comparing relative to my prior discharge events.)
To me it seems that charging event energy and discharging event energies both point towards the BMS rated miles estimate as being an accurate assessment of energy available.
But, if that is true, I am curious about exactly what happened within your battery that made more energy available over time! 15-20 rated miles is quite significant!
Are you sure the 270 at 100% was not measured at a time when the battery had insufficient time to balance? I could see doing a charge to 100% and then immediately driving. This would lead potentially to a result where your 100% was limited by the voltage cap on your weakest brick (since it would max out soonest). If the battery had been left at that state, of course it would have bled down that brick and then started charging all bricks in series some more - which would have resulted in more energy available than 270 rated miles.
To me that’s the first idea that comes to mind for how available energy would go UP in a battery over time. I assume it is not self-healing! And we know capacity loss in general is real, as the battery has various deleterious physical mechanisms by which it can hold less energy over time.
So what was the exact situation on the charge to 100% (you said the 270 was not an extrapolation, I think - obviously we ignore any extrapolation data points for this discussion)? Did it sit overnight or for 10-24 hours? Or did you charge and then go?