I’ve had a used Tesla Model 3 for about 3 months now and I’ve noticed my percentage increases towards the middle of the battery. For example, if I am on a drive and I start my trip with 60% battery and end the trip with 45% battery and park my car, after a few hours I come back and my battery is back up to 50%. It only does this in the 45-59% range.
Also, I will charge my car every night to 80% and it will finish after only about an hour and a half or so. When I get in my car in the morning, it’s down to 76-77%. It seems like my BMS may be messed up, but I’ve tried parking my car with 20% or less for a few hours and then charging back up to 100% but this problem still persists. It’s not a big deal, just thought it was odd. Has this happened to anyone else?
Yes, this is normal. When discharging or charging, as far as we can tell, the car does "dead reckoning" to determine the SOC. It takes the original SOC, meters energy added or removed, and "dead reckons" the new SOC (based on its best guess of current pack capacity). Only after the car sleeps does the BMS adjust its estimate of the "actual" SOC based on the open-circuit voltage and whatever other measurements it does.
It makes sense that the accuracy of the SOC estimate will depend on the particular SOC, because the SOC vs. voltage curve is non-linear, even though the BMS & the dead reckoning likely try to account for this. But there are a lot of moving pieces. It's hard.
Most likely in the charging scenario you aren't "losing" energy (I'm assuming the car is sleeping nicely); it's just that it measured the added energy and its effect on SOC incorrectly (possibly because it doesn't know the actual pack capacity).
It's pretty complicated (and inscrutable!) and depends on vehicle type, vehicle model year, software updates from Tesla, etc.
If you had SMT (or even if you don't), and you did a full discharge and metered the use and compared to Nominal Full Pack, you'd probably find that your BMS is currently underestimating the pack capacity. Your two observations suggest that is what is happening (there's more energy left than it thinks after a discharge, and the SOC is lower than expected after energy is added - these suggest the pack is larger than it thinks it is). Eventually, in theory, the BMS estimate of your full pack capacity will increase (rated miles at 100% will increase) once the BMS is convinced it is wrong.
If you observe this over a few weeks you might see this happen. Or you might not, since the exact behavior may change on a future software update.
If you have an LFP pack, you should charge to 100% periodically as directed. This helps the BMS to understand what the true Nominal Full Pack is, and theoretically would reduce this sort of error, eventually. You didn't specify your vehicle. But anyway NFP errors can occur in NCA packs as well. And charging to 100% won't magically fix all problems.
