Which is worse? 100% SOC or nearly empty?

[David99]

Some evidence from other threads that beneficial battery balancing occurs at charge levels >90%, which implies the opportunity to balance the pack without having to go fully to 100%. Agree no verification from Tesla, but some of us did see range improve a few miles after several weeks of charging to 90% (vs. charging to lower levels). Maybe some day we'll really know.

Some people have reported a slight increase of their rated range display after charging beyond 90%. This can be explained in different ways. It might be balancing, it might be just recalibrating the capacity algorithm as Tesla officially said it would. If the Leaf balances it's batteries at any time (while driving, charging or just sitting), I would assume Tesla would be limiting it to just near 100% charging. The size of the resistors found in the Tesla battery are too small to do any significant balancing in a short amount of time, thus a more continuous approach would be more likely.

 

[stopcrazypp]

Agreed. When the LEAF first came out drivers claimed they 'rebalanced their packs' by charging to 100% and leaving it there for x number of hours. Neither Nissan nor Tesla have revealed how dynamic rebalancing is accomplished by the vehicle and there ceratinly is nota user procedure to invoke it. Add on meters have shown the LEAF pack continuously rebalances groups of cells as it drives/charges/sits still. My guess is Tesla do something similar.

The Roadster had a top balancing procedure that was confirmed (both by procedures done by technicians and by monitoring the charging after reaching 100%), so it is not blind speculation. Of course the Model S BMS design may have changed (there is no OVMS so harder to monitor).

And it's quite common to have a regular maintenance balance and a more accurate top balance for BMS. People have reported squeezing a couple more miles by letting the car charge a long time on the top (although to be fair, it can also be a result of battery meter recalibration rather than balancing).

Conventional wisdom says that more damage is done by running a battery very low on charge vs charging to 100%. I would choose 100% over running very low. You may also avoid a tow if you miscalculate slightly

Without a anti-bricking mechanism and given vampire drain, having a battery low is more dangerous because of the risk of bricking. Of course, the Model S does have a anti-bricking mechanism (and is warrantied against bricking), so this point is moot. There is no damage mechanism from a lithium ion battery sitting at low SOC (unlike other battery types, lithium ion batteries do not have a memory effect).

However, a lithium ion battery sitting at high SOC and high temperature has accelerated degradation because the same higher reactivity from that state that allows higher power, also accelerates degradation.

 

[sorka]

So it's interesting. After it hit 100% this morning, instead of saying charging complete, it said "recalculating remaining charge time" and continued to charge for a good 20 minute while at 100% before I finally decided to call it quits and disconnect and head on out. Does this mean it was trying to balance out the cells?

I averaged 302 wh / mile on the 200 mile trip to Atascadero. There was only one other Tesla charging when I got there and when I left.

 

[stopcrazypp]

So it's interesting. After it hit 100% this morning, instead of saying charging complete, it said "recalculating remaining charge time" and continued to charge for a good 20 minute while at 100% before I finally decided to call it quits and disconnect and head on out. Does this mean it was trying to balance out the cells?

That's exactly the same behavior that have people suspect balancing is going on.

In the Roadster, it is easy to confirm that such behavior is balancing because there is the OVMS available that can read detailed data on what is happening to the battery, but unfortunately we don't have that on the Model S (and now there is no longer easy access to the service menu, or it would be possible to tell by looking at changes to the modules in the battery screen).

 

[Andyw2100]

So it's interesting. After it hit 100% this morning, instead of saying charging complete, it said "recalculating remaining charge time" and continued to charge for a good 20 minute while at 100% before I finally decided to call it quits and disconnect and head on out. Does this mean it was trying to balance out the cells?

So it sounds like you wound up doing one of the things you were trying to avoid--charging to 100% without getting the full benefit of pack balancing.

You may be interested in this thread that I had started when I was going to range charge for the first time, and wanted to make sure to leave enough time to let the pack balancing take place. I received some pretty good information. Question concerning timing the end of my first 100% charge

 

[sorka]

So it sounds like you wound up doing one of the things you were trying to avoid--charging to 100% without getting the full benefit of pack balancing.

You may be interested in this thread that I had started when I was going to range charge for the first time, and wanted to make sure to leave enough time to let the pack balancing take place. I received some pretty good information. Question concerning timing the end of my first 100% charge

I might agree had it not been for the fact that I let it do that for 30 minutes after it hit 100%. I suspect it must have been close but I needed to leave. I also didn't want to leave it at 100% for longer than necessary. How long could it have taken to rebalance? Could that have gone on for 45 minutes? An hour? 2 hours?

BTW, after that, my rated range at 80% went from 205 (which is more than the EPA rated range) to 203 which is 80% of 254. Assuming I didn't lose 2 rated miles but that the calibration drifted to the overestimating side and is now correct.

 

[jerry33]

I might agree had it not been for the fact that I let it do that for 30 minutes after it hit 100%. I suspect it must have been close but I needed to leave. I also didn't want to leave it at 100% for longer than necessary. How long could it have taken to rebalance? Could that have gone on for 45 minutes? An hour? 2 hours?

Some have had it longer than that. However, because it's still charging, it's not a problem. What you don't want to do is to charge to 100% and leave it sit for a couple of days--especially when hot.

 

[sorka]

Some have had it longer than that. However, because it's still charging, it's not a problem. What you don't want to do is to charge to 100% and leave it sit for a couple of days--especially when hot.

Sure. In this case I simply ran out of time. I'll know better next time that if I charge to 100% to time it so that it hits 100% a few hours before I leave. Hopefully it did the bulk of rebalancing before I actually unplugged it.


What benefit does rebalancing actually do in terms of battery health? If a battery in a module is out of balance and finishes charging first, will it heat up more than the other batteries that have a lower SOC and then spill over into the lower SOC batteries or will the extra charge simply spill over to the lower SOC batteries without the 100% full cell heating up?

Is the real danger doing a full discharge to 0% that lower SOC modules in a series could go to negative voltage?

- - - Updated - - -

OK, here's another one.

My commute is 128 miles each way. There is a super charger at the halfway mark. If I charge to 85% and drive to work, I'll have 35% when I arrive and then charge at work in the early evening after most of the EV spots have cleared out(to avoid door dings).

So I'm basically using 50% of the capacity of the battery going each way and I keep it in a range from 35% to 85% most of the time.

Would it be better to stop at the supercharger at 60% and charge back up to 85% so that I have 60% when I arrive at work and then charge back up to 85% in the evening while at work?

This would keep the range from varying no more than 25% but it would mean using the supercharger much more than I do now.

Would some other range be better, say 30% to 80%? If I started at 80% rather than 85% I guess the half way supercharger stop would charge a little faster as well since I'd start with a lower SOC.

 

[Todd Burch]

If the cells are not balanced, the cells with less charge get stressed more for a given usage pattern. They have lower voltage so their discharge is deeper, plus they experience greater stress than higher voltage cells when you floor it.

So having the pack balanced ensures even wear and tear on the cells.

Regarding 35% vs. 60%: 35% is probably better, but if you look at a graph showing degredation at these two values, the difference is so small that it's not worth worrying about. Anywhere less than 90% and above about 10%, try not to overthink it. I know it's hard, but just try to let the car do its thing.

 

[Andyw2100]

There was a post in the thread I referenced up thread a couple of days ago that adds some interesting information, I think, to what we know about what happens when we charge to 100%.

Question concerning timing the end of my first 100% charge - Page 3

We just need some people who use Visibile Tesla and who very regularly charge to 100% to weigh in with some data to add confirmation to the pattern that we seem to have identified. It appears that once the 100% charge has completed, something--either data processing or some other aspect of pack balancing--continues for some time, and that the amount of time may be dependent on how long it has been or how much the car has been driven since the last 100% charge.

 

[jerry33]

, here's another one.

My commute is 128 miles each way. There is a super charger at the halfway mark. If I charge to 85% and drive to work, I'll have 35% when I arrive and then charge at work in the early evening after most of the EV spots have cleared out(to avoid door dings).

Todd answered the balancing question. For the rest, here's the best write up I've found.

Originally Posted by hcsharp
The cathode in Li ion batteries forms defects (microcracks) due to stress caused by expansion and contraction relative to neighboring materials. It is simply bigger or smaller depending on whether it's charged or discharged. The more you discharge it, the more it changes size. And the more it changes size, the more microcracks it gets. These microscopic cracks lower the battery's capacity. That's one reason why smaller cycles, more often, contribute to longer battery life. That's why you should plug it in every night.

Heat aggravates the microcracking process, so keeping your battery cool contributes to longer battery life. And guess what? Your battery heats up more when used at a lower SOC because it requires more amps to keep your car going 65mph than it does at a higher SOC. That's why you should charge it every night.

Capacity fade also comes from the build-up of non-soluble deposits on the anode and cathode. This chemical process happens faster when the battery is warmer. It also happens faster when at a high SOC. But the process slows to a crawl when you drop the SOC to 80 or 90%, and slows only a tiny bit more at 50%. So if you are going to drive your car, keeping it charged in std mode has less impact on battery life (lower amps, less heat) than driving at a lower SOC. But if you're not going to drive your car for a few days, there are no amps or heat to worry about. That's when Tesla recommends putting it in storage mode, which keeps it at a lower SOC.

end of science lesson.

 

[sorka]

So the real question is, is hitting it with a super charger half way on the trip going to cause it to heat such that it's more detrimental than letting the SOC drop to 35% during the end of the journey?

If the SOC is 55% will the same cracking occur while the battery is subjected to high charging right or will it not care since it's not near the extremes of the charge range?

I'm wondering if it would be better to charge to 90% every night so I have 40% when I arrive. Yes I know we're talking minor differences here and I'm not really worried about it rather than just interested in discussing it. If makes even a small difference, then it's easy enough for me to raise my daily charge limit from 85 to 90.

 

[Todd Burch]

There was a post in the thread I referenced up thread a couple of days ago that adds some interesting information, I think, to what we know about what happens when we charge to 100%.

Question concerning timing the end of my first 100% charge - Page 3

We just need some people who use Visibile Tesla and who very regularly charge to 100% to weigh in with some data to add confirmation to the pattern that we seem to have identified. It appears that once the 100% charge has completed, something--either data processing or some other aspect of pack balancing--continues for some time, and that the amount of time may be dependent on how long it has been or how much the car has been driven since the last 100% charge.

This is pretty well-known behavior for us "old-timers". That's precisely why there's a belief that the car is balancing the cells once you hit 100%...also a reason why you should periodically charge to 100%. I heard someone secondhand say that a Tesla employee recommended once a month. All conjecture of course.

- - - Updated - - -

Sorka,

Tesla's spent a lot of time developing charging algorithms that maintain optimal charging power and pack temperature. It really doesn't matter, since they ramp down charging power as state of charge increases.

Nobody would probably ever notice the difference between 35% and 50%.

 

[jerry33]

So the real question is, is hitting it with a super charger half way on the trip going to cause it to heat such that it's more detrimental than letting the SOC drop to 35% during the end of the journey?

Unlikely because the rate of charge slows as the SOC increases. The main reason for letting the battery go fairly low before Supercharging is to speed the charge assuming you don't need most of your range. So total trip time with the battery going from 30% to 70% is slightly faster than from 50% to 90% (assumes the same driving speed).

 

[Cottonwood]

Unlikely because the rate of charge slows as the SOC increases. The main reason for letting the battery go fairly low before Supercharging is to speed the charge assuming you don't need most of your range. So total trip time with the battery going from 30% to 70% is slightly faster than from 50% to 90% (assumes the same driving speed).

And Supercharging from 5% to 45% is even faster. I wonder how much of a difference going for shortest charging time by planing lowest arrival SoC at a Supercharger, makes?

 

[sorka]

And Supercharging from 5% to 45% is even faster. I wonder how much of a difference going for shortest charging time by planing lowest arrival SoC at a Supercharger, makes?

39% = 112KW

59% = 74KW
60% = 72KW
61% = 70KW
62% = 68KW
63% = 66KW
64% = 65KW
65% = 64KW
67% = 60KW
68% = 59KW
70% = 58KW
71% = 56KW
72% = 55KW
73% = 54KW
74% = 53KW
77% = 48KW
83% = 40KW

So it make a huge difference. Below 35%, I think it's pretty flat unless they come up with super chargers that charge above 120KW in the future. When I plugged in this weekend at 20%, I think it was only like 115KW.

 

[Andyw2100]

This is pretty well-known behavior for us "old-timers". That's precisely why there's a belief that the car is balancing the cells once you hit 100%...also a reason why you should periodically charge to 100%. I heard someone secondhand say that a Tesla employee recommended once a month. All conjecture of course.

I had read a lot about the time spent before the charge actually finished, while the car was showing nearly 100%, or even 100% but still charging.

What I thought was new information, because I've read quite a bit, and don't believe I have seen it mentioned before, is that the range changes several hours after the charge actually completes, and that the amount of time before the range changes very well may be based on how unbalanced the pack was.

Again, this is different from the extra time to get, say from 99% to "Charge Complete." I know that that has been well documented. I'm talking specifically about the range increasing hours after the "Charge Complete" message. If that has been talked about before, I definitely did not see it.

 

[TexasEV]

I had read a lot about the time spent before the charge actually finished, while the car was showing nearly 100%, or even 100% but still charging.

What I thought was new information, because I've read quite a bit, and don't believe I have seen it mentioned before, is that the range changes several hours after the charge actually completes, and that the amount of time before the range changes very well may be based on how unbalanced the pack was.

Again, this is different from the extra time to get, say from 99% to "Charge Complete." I know that that has been well documented. I'm talking specifically about the range increasing hours after the "Charge Complete" message. If that has been talked about before, I definitely did not see it.

The displayed range is an estimate based on algorithms that no one knows other than we know Tesla has changed it several times. It may be a more accurate estimate that you're seeing later, but the actual range doesn't change.

 

[Andyw2100]

The displayed range is an estimate based on algorithms that no one knows other than we know Tesla has changed it several times. It may be a more accurate estimate that you're seeing later, but the actual range doesn't change.

Whatever. Whether or not the real range has changed isn't the point.

My point is that until now I don't believe anyone has identified the fact that long after the car reports "Charge Completed" the car is still computing and/or doing something, such that hours later it will report a different range, and that the length of time it takes to complete whatever it is that it is doing appears to possibly be be based on how unbalanced the pack was.

 

[andrewket]

Whatever. Whether or not the real range has changed isn't the point.

My point is that until now I don't believe anyone has identified the fact that long after the car reports "Charge Completed" the car is still computing and/or doing something, such that hours later it will report a different range, and that the length of time it takes to complete whatever it is that it is doing appears to possibly be be based on how unbalanced the pack was.

Perhaps not in such specific terms, but it was observed a long time ago (years) that rated range could go up a few hours after charging completed. I think people have even captured this in visible tesla; the graph does an instantaneous climb.