Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Percentage battery charge is a percentage of what? (2021 MIC M3LR)

This site may earn commission on affiliate links.
I have a Model S so may not apply to the MIC M3LR but the following has given consistent results over 5+ years and multiple software updates:

Without using 3rd party apps I can get two measures of capacity: one for charging and one for discharging.

For charging capacity, as suggested above, if you can measure (meter) how much kWh is needed to add, say, 50%, to the battery SOC (state of charge) you can extrapolate to 100% to get the energy needed to fill the battery. Doing this over several charge sessions and taking an average/doing a line fit on a spreadsheet would give a more accurate estimate. If AC charging the current (10A / 16A / 32A) makes a noticeable difference with the highest current being most efficient.

For discharge, if you note the % before setting off on a long, non-stop (i.e. don't leave the driver's seat so the car doesn't power down) trip the "since hh:mm" trip info can be used. E.g. depart at at 80%, arriving at 30%, 100 miles covered, 320 Wh/mi. Total energy used is 100*320/1000 = 32 kWh (so useable battery capacity is 64 kWh). If you set off immediately after charging then the "Since hh:mm" and "Since last charge" stats will tally and confirm the total kWh used. Again multiple long trips allow a more accurate estimate to be made - they should all plot on a straight line (energy vs. change in SOC) (if there is no measurable battery degradation over the period).

Finally as a way of measuring degradation look at the battery range in miles vs %. E.g. miles at 90% will gradually decline over time as the battery loses capacity. Rated miles corresponds to actual battery capacity. In my case (MS 70) 90% = 191 mi whereas when it was new 90% = 202 mi. You can calibrate the rated miles to kWh using the trip stats: for my car it works out at 1 rated mile = 287 Wh. This has stayed fixed over multiple software updates and the changing seasons.

I recall there may have been a bit of headroom above 100% on brand new cars (i.e. it was possible to add a bit more energy even when the SOC = 100) which helped to hide some initial degradation, but unless you charge to brim full you're unlikely to be affected by that.
 
If you want to keep it simple try TeslaFi. You don't give them your password, you only give then a token to access the API.
Also according to TeslaFi the average efficiency of a charge (240V @ 32A) is 91.44 % for my car over the last 3,000 miles
 
If you want to keep it simple try TeslaFi. You don't give them your password, you only give then a token to access the API.
Also according to TeslaFi the average efficiency of a charge (240V @ 32A) is 91.44 % for my car over the last 3,000 miles
this is a bit of a false sense of "security" to be honest, (unless something has changed recently that I'm not aware of) the token still gives access to the real(ish)time location of the car, ability to unlock, remotely start the car etc.

Who is behind teslafi, do they have any 3rd party attestations to back up their claims on how securely they handle your tokens other than "we'll pinky-promise to be really careful" ?. In the event of a breach, does your average user know how they could go about immediately invalidating a token?

Sure, in this day and age given all the cloud-security issues we see, it's not beyond the realms of possibility that Tesla themselves could suffer a problem that leaves your car in the hands of others, but in that event there would at least be some chance of recourse back to them (and they have the financial position to handle it) but if it was some random 3rd party website/app, I doubt an insurance company would be too interested in helping out.

If Tesla provided the ability for a read-only token, or an option to limit the commands a token can authorise (i.e. control charging & climate but not unlock & remote start) that would make things a bit better, but for now giving your tokens to someone else is a pretty bad idea.
 
Who is behind teslafi, do they have any 3rd party attestations to back up their claims on how securely they handle your tokens other than "we'll pinky-promise to be really careful" ?. In the event of a breach, does your average user know how they could go about immediately invalidating a token?

Quite frankly, TeslaFi has been around for ages with more than 15000 vehicles in the fleet. If something went wrong, it will be quickly picked up and requires a password change to lock out any further access to the car. I personally have the highest regard for James at TeslaFi, I have dealt with him personally on several occasions and would much rather let him have use of my token than I would an electricity company that many people have given the same access too. I'm not paranoid and have worked in IT and security long enough to know when there is too much risk involved.

My personal take is that it is better to let someone who knows what they are doing and has the check and balances in place to handle security 24/7 than letting someone who is not keeping an eye on things and keeping abreast of latest concerns 24/7 run a DIY setup. I am sure that if someone thinks they are capable of setting up a DIY solution, the are equally capable of forcing a token to be revoked.
 
Quite frankly, TeslaFi has been around for ages with more than 15000 vehicles in the fleet. If something went wrong, it will be quickly picked up and requires a password change to lock out any further access to the car. I personally have the highest regard for James at TeslaFi, I have dealt with him personally on several occasions and would much rather let him have use of my token than I would an electricity company that many people have given the same access too. I'm not paranoid and have worked in IT and security long enough to know when there is too much risk involved.

My personal take is that it is better to let someone who knows what they are doing and has the check and balances in place to handle security 24/7 than letting someone who is not keeping an eye on things and keeping abreast of latest concerns 24/7 run a DIY setup. I am sure that if someone thinks they are capable of setting up a DIY solution, the are equally capable of forcing a token to be revoked.

Of course, people are free to give their tesla tokens to whoever they like, but my point stands... it doesn't come without risk and just because 14999 other people are doing it, that doesn't mean it's a good idea.
 
  • Like
Reactions: JonB
I like the idea of a self hosted solution, it means I control it, not a 3rd party. TeslaFi is just another attack surface (in infosec speak).

But, let’s ask a question here: what does TeslaFi give you that TeslaMate doesn’t? They both use the same APIs, surely.

@init6 I think the answer to your question is that it just pulls data from Tesla’s servers via an API that you need to provide credentials to access, that is already being collected from your vehicle. The amount of info this entails is a bit scary…

The discussion is drifting off topic by the way, but it is very interesting.
 
I like the idea of a self hosted solution, it means I control it, not a 3rd party. TeslaFi is just another attack surface (in infosec speak).

But, let’s ask a question here: what does TeslaFi give you that TeslaMate doesn’t? They both use the same APIs, surely.

@init6 I think the answer to your question is that it just pulls data from Tesla’s servers via an API that you need to provide credentials to access, that is already being collected from your vehicle. The amount of info this entails is a bit scary…
I've never actually used it (for reasons above) but when I originally looked into it, TeslaFi does have some pros, like "push" commands, such as the ability to schedule climate & charges etc and whilst Tesla are catching up with their own 1st party support in the more recent updates of the mobile app, I think the general consensus is that the Tesla UX for this stuff isn't great at the moment and the logic can be a bit confusing. TeslaMate also gives you comparisons against their "fleet" so you can see how your battery is holding up compared to others etc.

TeslaMate is only pulling data, no commands can be sent (even though it needs the same tokens), the pros are as you said smaller attack surface (likely one or two vehicles behind any given instance, not tens of thousands) and the fact that it doesn't need to be publicly accessible. All the data is yours and stays with you. Plus if you're already familiar with SQL and Grafana you can extend the dashboards to your own liking.

It really depends on what functionality you want/need. For me, it's just a data-logger, the simple fact that there is an API to consume in the first place makes the whole ownership experience a bit more fun and interesting (I not sure if it's important to know that the last trip to waitrose cost 9p and was 48 seconds slower than before, but it's fun to know that info is there). Also helps with the mileage expense claims (when we're not working from home.

There are plenty of others (SaaS and self-hosted) but it seems these 2 are among the most popular, and from what I've seen on here and reddit, fans of each can be a bit "tribal".
 
@mjmherts I just got a text from "AnyVan" with my delivery time tomorrow, between 10am and 2pm!

How does Teslafi measure the amount of electricity you have used? Does it get data from the charger?
kWh added is obtained directly from the TeslaFi API. It is calculated by the difference between the starting energy_added field and the ending energy_added field of the charging session.

kWh used by the charger is calculated by TeslaFi and not directly obtainable by the API. Each data point is calculated to obtain the kWh by using ((Amps * Voltage) / 1000) * Phases. If a connection problem occurs and a data point is missing, the missing point will be replaced with the last good data point. A charge that occurs with more than 10 missing data points will be indicated at the charging session that the data may be inaccurate.

TeslaFi security. I guess if hacked it might be possible to unlock the car but not drive it away if the car is protected by pin to drive... And for additional security, and if you are worried, you can also limit the control TeslaFi has over the car in its settings, making unlock, etc unavailable...
 
This "lie" accusation is just hyperbolic nonsense. The WLTP range is the only official range that can legally be advertised by Tesla. Of course it's rare to achieve in the real world but nevertheless it is done from time to time.

I agree with all of this, but why do people say that some other EVs car makes (Eg Porsche) seem to more easily achieve or exceed their WLTP range compared to Tesla’s quoted range.

Is this even true? Are some car’s WLTP range more accurate than others?

I’m not complaining about the real world range of 280 miles I get in my 2021 M3 LR (same model as OP).

But I do get that some people may take issue with Tesla’s quoted range which seems particularly optimistic.
 
I agree with all of this, but why do people say that some other EVs car makes (Eg Porsche) seem to more easily achieve or exceed their WLTP range compared to Tesla’s quoted range.

Is this even true? Are some car’s WLTP range more accurate than others?

I’m not complaining about the real world range of 280 miles I get in my 2021 M3 LR (same model as OP).

But I do get that some people may take issue with Tesla’s quoted range which seems particularly optimistic.

I don't know the answer to this for sure ... but I would suggest that Tesla vehicles achieve their ranges through vehicles that achieve their potential range by virtue of higher efficiency through optimisation of aero, motors, weight reduction, battery design, cooling/heating, battery management etc etc whilst some other vehicles achieve their ranges more through having a larger capacity battery, but with overall less efficiency. This means that when everything is optimal the Tesla will take full advantage and achieve its rated range. The other side of the coin is that there is the potential for the Tesla to have a wider range of ranges (!) when factors that negate these optimised efficiencies come into play e.g. weather variations. The car with lower efficiency but a bigger battery perhaps achieves its range more by "brute force".

(It is a bit like where an ICE vehicle is so inefficient in its use of energy during the summer (throwing excess heat away) that during the winter you don't notice the mileage hit as much because it just throws away a bit less heat into the atmosphere and uses it to keep systems and cabin warm "for free".)
 
  • Like
Reactions: MrBadger and JonB
kWh used by the charger is calculated by TeslaFi and not directly obtainable by the API. Each data point is calculated to obtain the kWh by using ((Amps * Voltage) / 1000) * Phases. If a connection problem occurs and a data point is missing, the missing point will be replaced with the last good data point. A charge that occurs with more than 10 missing data points will be indicated at the charging session that the data may be inaccurate.
That's the same calculation that Teslamate uses. It's not as accurate as measuring at the charger. For example my last charge on a public charger consumed 38.17kWh, using the above calculation it registered 34kWh.
I don't care though as it's free :)