State of Charge Data over Time: what SoC is your battery mostly sitting at?

[sillydriver]

Well, if nothing else, you may then want to try 80 amp charging when it is cold, and you need to heat the battery, but in the spring and summer, when that's not an issue, consider dropping down to 56 amps.

That makes sense.

 

[MarcG]

What happens in the individual cells is a chemical reaction. No 2 cells, and a 85 kWh battery contains 7104 of them, are 100% the same. So after some charge/decharge cycles, they get slightly "out of balance" = they do no longer all sit at exactly the same voltage. This reduces the overall performance characteristics of the battery. When driving the car up to a very low SoC, the cells sitting at the lowest voltage might experience more "stress", which impacts their useful lifetime. To keep your battery pack in the best condition, the individual cells need to be "balanced" every once in a while. Thanks to the research of our much respected forum member wk057, we know that the balancing mechanism is triggered when the battery is charged above around 93% SoC. So this is essentially when you do a range charge. You don't have to do anything special to trigger balancing, except do a range charge e.g. once a month or once every quarter. It will make sure all cells are calibrated so they are all at the same voltage level again.

Since I keep my cells at a low SoC most of the time, the battery management system has a lot of trouble measuring / keeping track of capacity. I know this because the range displayed in my car decreases faster than the range displayed in cars that are charged to e.g. 90% on a daily basis. This is because it is very difficult to keep track of a cell's capacity if it is not completely discharged and charged. So in my case balancing regularly is even more important to make sure individual cells don't get too far out of balance with the rest. This could result in over discharging or over charging these individual cells which is very bad for their health.

Great explanation, thanks!

So in your case, after you do a range charge, do your subsequent 90% charges restore some of the "lost" miles?

 

[apacheguy]

Since I keep my cells at a low SoC most of the time, the battery management system has a lot of trouble measuring / keeping track of capacity. I know this because the range displayed in my car decreases faster than the range displayed in cars that are charged to e.g. 90% on a daily basis. This is because it is very difficult to keep track of a cell's capacity if it is not completely discharged and charged. So in my case balancing regularly is even more important to make sure individual cells don't get too far out of balance with the rest. This could result in over discharging or over charging these individual cells which is very bad for their health.

My problem with this is that even someone who deeply discharges, say 70%->30%, on a daily basis sees more apparent loss than folks who discharge 90->75% daily. It's more than just the depth of discharge that Tesla uses to calculate capacity.

 

[ArtInCT]

My problem with this is that even someone who deeply discharges, say 70%->30%, on a daily basis sees more apparent loss than folks who discharge 90->75% daily. It's more than just the depth of discharge that Tesla uses to calculate capacity.

Apacheguy:

Can you share with us the algorithm that Tesla uses to calculate capacity? I would be most interested in learning this. Especially the weather effect (temp) and if that is mostly taken into account and logged by the Model S algorithm when the car is on the road and running or in a quiescent state, say parked overnight. Also is the algorithm more accurate if you are charging to a 90% State of Charge and if not, what is the penalty?

Also, is driving style a factor? Do many fast acceleration launches become part of a "profile" which is used by the algorithm.... on the flip side, if you hyperbole and drive as if an egg were between your foot and the go pedal, is that part of your "profile"? Is you driving style taken into account is the big question.

 

[tomas]

There are quite a few threads about this, plus charging FAQ. Loss of rated range has several causes. 1) calibration drift, 2) out of balance, 3) degradation. Most people who charge below 90 have found that discharge to <20% followed by charge to 90% or more, or repeated charge to 90%+ will recalibrate and gain miles. There's debate about exactly what level triggers cell balancing, but 100% charge certainly does. As far as degradation, there are some charts, but in summary it's not unusual to lose a few % first year, and <1% annually thereafter.

I personally agree with flasher. Other than avoiding leaving car sitting at 100 and leaving at 50 while sitting for > a few days, I don't worry.

i thought this thread started out very interesting: as an analysis of actual SOC experienced. It has swirled into a rehash of knowledge that is pretty well documented in loss of rated range thread. Not that that is bad, but curious readers have a place to go read more.

 

[apacheguy]

Can you share with us the algorithm that Tesla uses to calculate capacity? I would be most interested in learning this. Especially the weather effect (temp) and if that is mostly taken into account and logged by the Model S algorithm when the car is on the road and running or in a quiescent state, say parked overnight. Also is the algorithm more accurate if you are charging to a 90% State of Charge and if not, what is the penalty?

There have been links posted in other threads to patents filed on Tesla's behalf for this. I don't recall the specifics. Basically, it meters input and output and does some math to spit out CAC. How that number is translated to rated range is not clear.

Furthermore, there are enough threads on this to merit a blog post by Tesla. And it's getting to the point where new customers are wanting answers, particularly for the bizarre trend in 90 kWh packs. Why Tesla doesn't respond is not clear either.

 

[jpet]

So in your case, after you do a range charge, do your subsequent 90% charges restore some of the "lost" miles?

Yes, I don't do subsequent 90% charges but 70% charges :wink:, but I see the effect of balancing each time I do a range charge. Since I do not want to do a full cycle (starting from 0% SoC), I am not able to get my battery back to the original range at 100% SoC. I do not care about this because I know why my figures are worse than those of others. At some point I'll probably want or need to do a full cycle and then the BMS will be able to reset its parameters and restore the range figure to close to the battery's original capacity (minus some degradation of course).

By the way, it takes up to 1 hour and 45 minutes in my case to fully range charge my battery at a super charger when it is in need of balancing. I wait till the charging current drops to 0 Amps so I plan accordingly and order another drink or 2.

 

[tomas]

I suspect tesla doesn't respond because they've gone to great pains to design BMS to idiot-proof the system. They do not want owning BEV to be an arcane lesson in EE. Unfortunately, that leaves gEEks guessing, which I suspect many complain about but secretly love....

 

[apacheguy]

I suspect tesla doesn't respond because they've gone to great pains to design BMS to idiot-proof the system. They do not want owning BEV to be an arcane lesson in EE. Unfortunately, that leaves gEEks guessing, which I suspect many complain about but secretly love....

I suspect you're correct. But I, for one, would love to read a detailed explanation coming right from the source. The Model S is arguably the geekiest car on the road and should be supported as such.

 

[ArtInCT]

I know quiet a few well heeled professionals who own Teslas.... have no time to geek out... and just plug in ...and then.. in the morning GO.... no fuss no muss and when asked question like I have asked on this thread... they knit their brows... and become dismissive rather quickly. Recently, when I brought up the topic of balancing the battery with one of them, the answer was "Look, I have no time or inclination to do that at all..." Funny, this world of Tesla owners... for sure but certainly expected and normal... I on the other hand am probably NOT... normal that is

 

[sillydriver]

I know quiet a few well heeled professionals who own Teslas.... have no time to geek out... and just plug in ...and then.. in the morning GO.... no fuss no muss and when asked question like I have asked on this thread... they knit their brows... and become dismissive rather quickly. Recently, when I brought up the topic of balancing the battery with one of them, the answer was "Look, I have no time or inclination to do that at all..." Funny, this world of Tesla owners... for sure but certainly expected and normal... I on the other hand am probably NOT... normal that is

You're normal among the self-selected crowd of nerds participating in a forum about the ultimate nerdmobile, as am I.:wink:

 

[Old Man Mike]

If you dig deep in the previously noted Battery University website, you'll find this specific reference to Tesla's P85 battery:

Note that the Cycle life is 500 but is related to temp and depth of discharge. Based on this and my own experience with Li batteries, I'm pretty sure that the small recharges do not count significantly toward the referenced 500 cycle life. In fact, just think of the number of small recharges during brake regen. Tesla's recommendation to keep the car plugged in may be to reduce the number of deep discharge cycles.

As for battery degradation, I'm not a big fan of using calculated range as a metric. A measurement of Max Power during a 0-60 test for different SOC would provide a much better metric of battery condition. I will be maintaining weekly data of my P90D starting with the first day of delivery next week.

Finally, I've seen the posts suggesting that the newer silicon anode technology of the P90 may have faster degradation than the P85. Certainly the early inclusion of silicon resulted in lower life cycles. However, if you do a detail look at the development in recent years, the use of silicon nano fiber has life cycles 50% greater (750 Cycle life). Hard to know the exact implementation that Tesla is using but as a world leader in battery technology, I suspect P90 battery performance at least as good as P85.