Estimated Battery Degradation - Model 3

[Medved_77]

I re-read this thread last night:

How I Recovered Half of my Battery's Lost Capacity

TL;DR

BMS reports an estimated range value.
TM3 batteries can take up to 6 hours of resting to stabilise voltage to obtain a reading due to number of resistors.
Periodically charge to greater than >90% and let the car sleep for >6 hours.
Periodically let the car drain to <10% and let the car sleep for >6 hours

I don't understand the science well enough to know how/if this works but it was an interesting read.

 

[Durzel]

Interesting thread linked there @Medved_77

According to TeslaMate my car has lost circa 50 miles from it’s rated range, in less than 1000 miles...

I’m not bothered though, because a) if I actually needed to go a long distance I’m sure it would go further than it currently thinks and b) the battery has an 8 year warranty. If it actually did become a objectively real problem I can deal with it then.

I suspect in my case it’s because I barely do any miles at all at the moment, and the journeys I do are short, so the data is mostly useless.

If I just took it at face value though I ought to be concerned.

 

[Medved_77]

Interesting thread linked there @Medved_77

According to TeslaMate my car has lost circa 50 miles from it’s rated range, in less than 1000 miles...

View attachment 611377

I’m not bothered though, because a) if I actually needed to go a long distance I’m sure it would go further than it currently thinks and b) the battery has an 8 year warranty. If it actually did become a objectively real problem I can deal with it then.

I suspect in my case it’s because I barely do any miles at all at the moment, and the journeys I do are short, so the data is mostly useless.

If I just took it at face value though I ought to be concerned.

Firstly, I can't believe you've only done 771 miles since March!!!

I don't use the car for commuting yet have been able to fit in just under 5000.

Looks like your BMS is reporting a 30% range loss where mine is closer to 50%:

The key thing I took from the article is that the physical battery is most likely fine, it's just the calculated estimate range that needs tuning now and again. I'm going to follow the advice during the next good Agile period and see if it makes any difference.

 

[Glan gluaisne]

I suspect in my case it’s because I barely do any miles at all at the moment, and the journeys I do are short, so the data is mostly useless.

I think you're spot on with this conclusion, and believe that far too many people are reading very approximate estimates of predicted range as if they are based on lots of reliable data, when the reality is they almost certainly aren't.

I've been playing around with electric vehicles for around 20 years, have built several electric bikes, an electric motorcycle and a solar powered electric river boat. In that time one thing I've learned about lithium chemistry batteries is that it's near-impossible to estimate remaining capacity to better than about 10% to 15%, even when you have loads of measured data. My boat has a pretty comprehensive measurement system, that logs both charge and discharge current (and voltage) as well as individual cell voltages and temperature.

The latter has far and away the greatest impact on apparent state of charge, due, I suspect, to mechanical changes associated with variations in cell internal pressure, and hence electrode spacing. Lithium cells seem very sensitive to the latter, back when I was using pouch cells to build packs I found that carefully controlling the pack preload pressure on the cells was key to getting consistent performance (and making cells last longer).

I'm sure the BMS is measuring as much data as is physically practical, some of which can be accessed via the API, but with the best will in the world I don't think it will be able to get better than maybe a 10% approximation as to remaining capacity at any time. Compared to the accuracy of a fuel gauge on a conventional car that's probably somewhat better, overall, and so was probably deemed to be good enough. I doubt that Tesla thought that people would then devise ways of extract data from the BMS, using tools like Teslamate, etc, and then trying to use the car's measurements and estimates for other purposes.

 

[Durzel]

Firstly, I can't believe you've only done 771 miles since March!!!

I don't use the car for commuting yet have been able to fit in just under 5000.

Looks like your BMS is reporting a 30% range loss where mine is closer to 50%:

View attachment 611378

The key thing I took from the article is that the physical battery is most likely fine, it's just the calculated estimate range that needs tuning now and again. I'm going to follow the advice during the next good Agile period and see if it makes any difference.

I know. When I ordered the car I was doing a minimum of 200 miles a week commuting. That went to 0 with lockdown, and then subsequently working from home full time.

The sad part is that I could have probably hired a car between then and now, and bought a refreshed Model 3 in the new year (when I'm expecting the company I work for will start making overtures about going back in) and saved more money than I've lost in depreciation, probably. Still, I love the car

 

[Jeeves]

The key thing I took from the article is that the physical battery is most likely fine, it's just the calculated estimate range that needs tuning now and again. I'm going to follow the advice during the next good Agile period and see if it makes any difference.

I think that’s the right takeaway.

Whilst I can understand anyone wanting to be reassured that their battery degradation is normal, I think we have too much data available that is often misleading. I was initially fascinated by it all, now I rarely bother looking beyond Wh/mi and whether or not the car thinks it can get me to the next charger within 2–3 hours. The answer to the latter is almost always: easily.

I’ll still be interested to learn of the result of your experiment though!

 

[hgmichna]

My plan is to charge the car as per norm (daily charge at 90%) and around January do another 100% and see what i get.

To stretch the lifetime of your battery just a little bit more, you could consider charging only to 80% (or even less) when you are sure you don't need more the next day.

 

[AnthonyLR]

Elon said yesterday Tesla batteries are currently designed to last 15 years. Naturally there will be some degradation but interesting to hear him say that.

 

[Drew57]

...Started last week & currently part way through (@ 42%) running the batteries down to 10-15%. Normally charge 80-90% but hardly used the car for a couple of weeks apart from daily tip runs while clearing the loft during lockdown. Last charge was over a week ago.

In 5000 miles I've rarely been below 50% & a lot of time 70-90% so thought it best to give the BMS a chance to balance.

 

[Durzel]

I plugged my car in on Monday, just to see what would happen (usually I've not bothered to leave it plugged in at home, as I'm not driving anywhere).

Some observations:

• It's woken up at the scheduled charging time (3:30am) every morning, even though my charge limit is at 50% and it's at 56% charge. Would be interesting to know whether it will wake up every single time even if there's no practical way it could lose the difference between current charge and charge limit (e.g if it was charged to 90% when it last fell asleep).
  
  
• Whereas before it would stay asleep for days, the last two days it has spontaneously woken up at 16:00, before going back to sleep again. No idea why.
  
  
• On more than one occasion after waking up and falling asleep, it has ended up "offline" in TeslaMate. Not sure why this is happening. I have been opening the app which wakes the car up, and then when it falls asleep again it shows "Asleep" correctly. Happened today at the scheduled charging time (03:30) and also at 16:00 (the spontaneous wake time). Strange.
  
  
  
  
• It's not drawn any power at all from the chargepoint, at any time.

 

[CMc1]

Some comparative data here. Please add data Tesla Battery Survey

 

[Durzel]

Thought I'd bump this. Although it's not been a full 2 weeks since I plugged in the car and left it plugged in, since last Monday it has not used any electricity from the chargepoint whatsoever.

The car is currently at 52%, it's losing maybe 0.5% a day, possibly less. My charge limit is 50% when I'm not using the car, so I would presume if I left it another week it would start intermittently charging back up to that level.

From what I've experienced the car doesn't do any obvious "battery maintenance" from the mains, certainly none that is evident over the course of almost 2 weeks.

 

[CMc1]

Thought I'd bump this. Although it's not been a full 2 weeks since I plugged in the car and left it plugged in, since last Monday it has not used any electricity from the chargepoint whatsoever.

The car is currently at 52%, it's losing maybe 0.5% a day, possibly less. My charge limit is 50% when I'm not using the car, so I would presume if I left it another week it would start intermittently charging back up to that level.

From what I've experienced the car doesn't do any obvious "battery maintenance" from the mains, certainly none that is evident over the course of almost 2 weeks.

Is it supposed to occur when connected or unplugged? Can’t remember what they thread referred to

 

[Glan gluaisne]

I've only spotted my car doing a 2 hour turn on twice. Both times were after it had been parked up for a week or so, and I think both times were when it was cold overnight. Normally it turns on to "phone home" for about ten minutes a day, the longer periods occurred at the same time as the "phone home" events, but the car just stayed on longer. I'd assumed this was just topping off the 12 V battery, but have no easy way of knowing if that's really the case.

 

[CMc1]

Wonder if the recently released service manuals and diagnostic tools shed any light on how the BMS operates

 

[Medved_77]

Wonder if the recently released service manuals and diagnostic tools shed any light on how the BMS operates

Did you manage to gain access to those?

I registered my Chinese account, got access to the training programs and requirements for authorised dealers but my login for toolbox.tesla.com was not valid.

 

[CMc1]

Did you manage to gain access to those?

Haven’t attempted yet..

 

[Durzel]

Wonder if the recently released service manuals and diagnostic tools shed any light on how the BMS operates

My belief is that Tesla advise that the car is always plugged in simply because it eliminates the risk of the battery depleting completely by mistake, and also because it's not really a bad idea to plug in whereever its convenient.

Because I don't really do many miles having the car plugged in when it's not doing anything at all doesn't really benefit me. I was hoping to see the car do something in terms of spontaneously drawing power from the chargepoint, but in my experience over 12 days, it hasn't charged at all - even at the scheduled charging time (it wakes up and goes back to sleep).

If there is some 12v maintenance going on it doesn't appear to need shore power, or it occurs at a longer interval than two weeks. Although Tesla state that the car ought to be plugged in, I've not seen any behaviour that justifies it. It has literally drawn no power from the charger.

 

[Drew57]

^ @Durzel exactly my experience too - when plugged in I don't see any power draw apart from timed charges.

The car also completed its battery drain to below 15% last week & now back to regular 80%+ but without any noticeable changes to prior.

 

[UrbanSplash]

Car is a year old in 2 weeks. I’m not doing many miles. Charge up to 80% maybe once a week. Never charged to 100%. Used supercharger 4 times in a year. I’m not getting worked up about it.

appreciate it’s cold: