What's the typical/average between what the car tells you it used vs what Teslafi says it put into the vehicle?
The car's displays tend to emphasize driving use, rather than charging consumption. The driving use data omits preconditioning and vampire drain, so it can be pretty far off. The driving data also doesn't distinguish between home charging and charging elsewhere. I'm pretty sure that the charging screen does show a total kWh put into the battery, but I rarely check that -- rarely enough that I'm not even 100% sure of what data it shows.
Also, TeslaFi reports two numbers for charging: The into-the-car kWh and the into-the-battery kWh. The latter is recorded by the car, and the general opinion I've seen on this forum is that it should be quite accurate; however, for billing purposes, that number is not the most relevant one. For billing, the into-the-car number is more important, since there are losses/inefficiencies associated with charging. As I understand it, TeslaFi computes an into-the-car kWh number based on a minute-by-minute polling of the instantaneous kW reading. Thus, this number is slightly imprecise because the kW reading will fluctuate slightly over time, and the polling may miss a brief spike in use. Worse, if there's a data transmission error, some data will be lost and the number could be off by much more. That said, I've compared the into-the-car data as reported by TeslaFi to what my JuiceNet-equipped EVSE reports, and they match pretty closely, but not perfectly, presumably because of the polling and because of losses between the EVSE and the car. (Some power does go into heating up the cable, for instance.) My comparison was based on just a few charging sessions, though, so I may have gotten lucky or missed some huge glitches that could, in principle, be common.
As an example, here are what JuiceNet and TeslaFi report for my four most recent charging sessions at home:
Code:
Date JuiceNet Use TeslaFi Use TeslaFi To Battery
12/11 14.5 14.4 13.41
12/10 9.6 9.5 9.32
12/8 12.3 12.2 11.94
12/7 13.8 13.3 12.68
FWIW, I noticed that the 12/7 session included two brief "spikes" in use after the main charge was complete. The first was shortly after the main session, and I think it was caused by the car recovering power lost to vampire drain shortly before it went to sleep. The second was at the very end of the session and was probably caused by my waking the car to unplug it; between waking and unplugging, it drew power for a few seconds. These spikes were picked up by JuiceNet, but not by TeslaFi, presumably because they were very brief and so didn't register on TeslaFi's polling. Even on that session, TeslaFi registered 96.4% of the power that JuiceNet said was used. If greater precision than that is required, then TeslaFi may not be the best choice; but if an underestimate of ~4% is OK, or if you and the landlord agree to an adjustment, then TeslaFi (or something else based on the Tesla API) should be fine.