Given the differences in capacity between the two, the effect should be zero unless you're talking about very low SOCs where the software is trying to preserve the traction battery.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Given the differences in capacity between the two, the effect should be zero unless you're talking about very low SOCs where the software is trying to preserve the traction battery.
smorgasbord
Active Member
This has been discussed before, in the Roadster threads. It's not linear.
Tesla's answer has been:
The bold emphasis is mine. Yes, Model S battery chemistry is different, but not different enough to invalidate this behavior.
Not charging means you're battery will often be warmer than if you were charging more frequently. That seems odd until you realize that the car is designed to actively cool the battery during charging. So, the battery temp after 30 minutes of charging will usually be lower - quite a bit lower. That's where the Tattler's "cool-down" feature comes into play.
I don't know if the Roadster battery surveys collect the time spent at various temps (it's in the vehicle logs), but it would be interesting to compare the discrepancies between batteries of similar age and mileage having different CAC values against the time spent at higher temps. Tomsax - are you here?
Tesla's answer has been:
The bold emphasis is mine. Yes, Model S battery chemistry is different, but not different enough to invalidate this behavior.
Not charging means you're battery will often be warmer than if you were charging more frequently. That seems odd until you realize that the car is designed to actively cool the battery during charging. So, the battery temp after 30 minutes of charging will usually be lower - quite a bit lower. That's where the Tattler's "cool-down" feature comes into play.
I don't know if the Roadster battery surveys collect the time spent at various temps (it's in the vehicle logs), but it would be interesting to compare the discrepancies between batteries of similar age and mileage having different CAC values against the time spent at higher temps. Tomsax - are you here?
Thanks for the links and the percentages. From your figures, it seems like I could probably stand to gain quite a bit of battery life by charging minimally. So, while I do occasionally drive ~140 miles in a day, on maybe 70-80% of my days I will drive very little (22 miles if to work and back, maybe 35 miles if I also drive to lunch... and I am apparently designated driver for my lunch group now that I have a Tesla
), which will mean that I will be sitting at a fairly high SOC quite a lot a lot of the time (charge to 188, drop to 180 from general "sitting around drain", -30 = 150 miles, which is still a ~72% SOC... although honestly I am having a hard time parsing that capacity fade rate graph (having a hard time picturing what a capacity fade rate of -.003/(day^.5) means exactly...I feel like despite the fact that I normally don't drive that much, that I might be being pretty hard on my battery. Going between 70-90% charge all the time (high SOC), with the 2-3 time a week drop to 30-50 miles (15-25% SOC). Hrm.
I really wish there was a way to do the "charge to X% by 9:00 am" scheduling :/ By starting to charge at midnight or 2:00 am, I give myself enough leeway in case I've driven down to ~30 miles range the day before, but on all the other days, I'm sitting at 90% for ~7-8 hours before I go to work.
To those saying "Tesla says 90% so I'd trust that", I'd, uh, counter(?) by saying that it may be to their benefit in terms of mass-market appeal if they told people not to worry about it and gavepeople a percentage that leads to medium battery degradation and still gives you a really large driving range, rather than scare people off with instructions to maintain battery life by doing a bunch of math every night before you go to sleep that leads to much better battery degradation.
stopcrazypp
Well-Known Member
Not true. The negatives of a low SOC are:
1) You have a higher risk of over-discharging the battery (AKA bricking like with the Roadster)
2) The battery needs to be discharged at a higher C-rate for the same power. The rate will be inversely proportional to the battery voltage. Given nominal voltage for the cells are 3.6V and the lower cut off voltage is 3.0V, this is only a 17% difference even in the case of indicated "0% SOC".
For #1, it can't happen in a Model S because the BMS will prevent that (Tesla says the car is not brick-able). Even at "0%" indicated SOC, the actual battery SOC is actually around 2% (for the Roadster).
For #2, Tesla already has a fairly aggressive power limiter that kicks in at lower SOCs (see the threads about taking the Model S to the track). And also from the Panasonic I posted the difference in degradation from higher C-rate (even at 2x as much and 3000 cycles) is much less than from storage at high SOC:
1I, 25 degrees C: 8.2%
1I, 50 degrees C: 10.7%
2I, 25 degrees C: 13.9%
2I, 50 degrees C: 13.9%
1I is 1C or 85kW for the 85kWh Model S, 2I is 2C or 170kW for the 85kWh Model S.
- - - Updated - - -
Yes, but if you look at the figures, there's diminishing returns as you go lower. You get the most benefit avoiding 100% SOC and less and less benefit as you go down. It's a matter of balancing utility of the car with the degradation, and 80-90% seems like a good compromise for most people. And also you have to consider that fact that there's a "race" between cycling degradation and storage degradation (it's not additive). So you will be losing battery capacity from cycling anyways.
Yes, it's difficult to parse. But what that figure is basically is the for the formula: capacity change = capacity fade rate * (days^.5). So say for example you want to figure out the loss after 8 years = 2920 days. You just get the capacity fade rate, say for example 100%SOC at 25C, which is -0.00448. Then you get capacity change = -0.00448*(2920^.5) = -0.242, which is means a loss of 24%. I basically just did that on an excel spreadsheet to get my numbers.
Does the Model S not already have that option? People have been doing timed charging, but I guess they may be doing that outside the car?
Well Tesla would give a number that will make the car last past it's warrantied 8 year life, and ~80-90%SOC already does that.
Until Tesla changes their written statement on charging for the Model S, all we really have to go by is what's in the Owners Guide or what they put into the Release Notes for the software updates. Since the release notes for the new charging slider feature does not make a statement about how using this feature benefits the battery, all we have officially is what's in the Owners Guide. Any statements by individual staff members are not official policy - and are often contradictory.
On page 15, the current Owners Guide states that Tesla strongly recommends leaving Model S plugged in when not in use. THIS MAXIMIZES THE LIFETIME OF THE BATTERY (see page 25).
On page 17, it states that using MAX RANGE frequently reduces battery life.
On page 25, it states that "The most important way to preserve the Battery is to LEAVE YOUR MODEL S PLUGGED IN when you're not using it. This is particularly important if you are not planning to drive your Model S for several weeks. When plugged in, Model S wakes up when needed to automatically maintain a charge level that maximizes the lifetime of the Battery. There is no advantage to waiting until the Battery's level is low before charging. In fact, the Battery performs best when charged regularly."
Though Page 25 does have some erroneous statements about the average battery discharge rate being only 1% per day - when most owners appear to see about 10 miles of range lost per day - which is higher than 1%.
Assuming that the Owners Guide is still correct - and then Tesla is recommending the car always be charged to 90%.
And if that's the case, the only reason why I would use the new charging slider feature would be to limit the amount of charge from a "for pay" charger - why pay at a high rate to charge more than is needed until you can get the car to home, a Supercharger or a Tesla store/SC???
However, if the charging slider feature has been provided because Tesla has decided the simple rules in the Owners Guide are not the best practice - then they should update the Owners Guide and notify all owners - in writing...
On page 15, the current Owners Guide states that Tesla strongly recommends leaving Model S plugged in when not in use. THIS MAXIMIZES THE LIFETIME OF THE BATTERY (see page 25).
On page 17, it states that using MAX RANGE frequently reduces battery life.
On page 25, it states that "The most important way to preserve the Battery is to LEAVE YOUR MODEL S PLUGGED IN when you're not using it. This is particularly important if you are not planning to drive your Model S for several weeks. When plugged in, Model S wakes up when needed to automatically maintain a charge level that maximizes the lifetime of the Battery. There is no advantage to waiting until the Battery's level is low before charging. In fact, the Battery performs best when charged regularly."
Though Page 25 does have some erroneous statements about the average battery discharge rate being only 1% per day - when most owners appear to see about 10 miles of range lost per day - which is higher than 1%.
Assuming that the Owners Guide is still correct - and then Tesla is recommending the car always be charged to 90%.
And if that's the case, the only reason why I would use the new charging slider feature would be to limit the amount of charge from a "for pay" charger - why pay at a high rate to charge more than is needed until you can get the car to home, a Supercharger or a Tesla store/SC???
However, if the charging slider feature has been provided because Tesla has decided the simple rules in the Owners Guide are not the best practice - then they should update the Owners Guide and notify all owners - in writing...
Since 4.5, I've adapted to an 80% charge limit, and I now charge at 220 at 10-15 AMPS. I figure this does several things: keeps me charging longer (pack at ideal temp longer), keeps me away from extremes of battery state (away from 10% and away from 90+%), and charging at lower amps (slower) is supposed to be better for cell balancing. I don't always reach 80% from when I start my scheduled charge, but that might be good too.
The Model S lets you time charge, as in, start charging at this time. It doesn't let you time the end of the charge. ;-) Most people who're doing that, or all, are probably timing their start based on their average usage and how long that amount takes to charge. That's basically what I did. Although now I'm worried that by charging on these hot days, while the sun's out, I'm hurting my battery. Argh. ;-(
Brit4864
Member
From what I understand, and I'm no expert by any means, software versions earlier than v4.1 had a "sleep mode" enabled which is why Tesla say in the manual the discharge rate is only 1%. I guess they haven't updated the manual post v4.1. The 10 mile "vampire loss" appears to be because the battery continues to power the computers all the time. Tesla claim to be working on a new sleep mode function to be released by the end of this year.
The Release Notes are a bit unclear...
...and wasn't it Elon who said we should set the slider to what we need--implying only as much as we need? I hear what you're saying about individual staff members, but I think a lot of people (well, at least me ;-) ) tend to listen a little more closely when Elon speaks. Perhaps not a great idea. ;-)
Ah, I see. Are those capacity fade rate numbers (e.g. -.00448 for 100% SOC at 25C) written out in the paper you linked (maybe I'm blind but I didn't see it), or did you eyeball it from the graph in the figure? Another thing I should take into account is what %charge a "full charge" is. From that page about the roadster batteries, it looks like tesla limited the roadster charging so that a "full 100% charge" was really only a 95% charge. If that's the case, then our standard "90%" charge is really more like a ~85% charge, and an "80%" charge is really more like a ~76% charge. In that case, I'm further down the diminishing returns curve than I thought.
You can tell it when to start charging, but not when you want it to try to hit the end of charging. So I charge at 2:00 am every night so that on the days I need to charge 150 miles I can get all my charging done before I go to work, but on the days I only need to charge 30 miles, it finishes charging at 3:00 am and sits around at a high SOC for 6 hours. It would be better if I could tell it I wanted it to finish at 8:00 am.
Yes, agreed; but I think for them 70% battery capacity counts as usable, while for me (because I am dumb and did not really plan very well), 70% is not (I would not be able to do my 140 mile trips at 70%). I just meant that it may "sound worse" to consumers if they could have 85% battery capacity after 8 years at the cost of lots of busywork every night than 70% battery capacity after 8 years without having to think about anything.
djp
Model 3 Performance
You can estimate the number of years it'll take to lose 20% capacity by eyeballing the y-axis for a specific SOC and temperature. Years = (0.20/rate)^2 / 365.
For example, here are the number of years it will take to lose 20% capacity while storing the battery in both a cool climate (20C or 68F) and a warm climate (30C or 86F):
As a rough estimate, the drops from 90% to 60% to 20% SOC each double the storage life of the battery, and the drop of 10C also doubles the life of the battery.
As you mentioned the Roadster protects the top 5% SOC and only charges to 95% on a Range charge and 83% on a Standard charge. The Model S probably does the same, so the actual SOC would be roughly 5% lower than the displayed SOC.
The Roadster owners manual goes into more detail on charging modes than the Model S manual. Storage mode on the Roadster is roughly 20% SOC.
The Model S 4.5 release notes had guidance on charging which suggests lowering the SOC if you're not using the full range of the car:
Personally I charge my Roadster to 50% during the week and Standard Mode when I need it on a weekend. This gives me more than enough range on weekdays and it's not a big deal to top it up on Friday night or Saturday morning for a weekend run.
For example, here are the number of years it will take to lose 20% capacity while storing the battery in both a cool climate (20C or 68F) and a warm climate (30C or 86F):
| SOC | Cool | Warm |
| 100% | 7 | 4 |
| 90% | 11 | 6 |
| 80% | 14 | 8 |
| 60% | 23 | 12 |
| 40% | 36 | 19 |
| 20% | 49 | 27 |
As a rough estimate, the drops from 90% to 60% to 20% SOC each double the storage life of the battery, and the drop of 10C also doubles the life of the battery.
As you mentioned the Roadster protects the top 5% SOC and only charges to 95% on a Range charge and 83% on a Standard charge. The Model S probably does the same, so the actual SOC would be roughly 5% lower than the displayed SOC.
The Roadster owners manual goes into more detail on charging modes than the Model S manual. Storage mode on the Roadster is roughly 20% SOC.
The Model S 4.5 release notes had guidance on charging which suggests lowering the SOC if you're not using the full range of the car:
Personally I charge my Roadster to 50% during the week and Standard Mode when I need it on a weekend. This gives me more than enough range on weekdays and it's not a big deal to top it up on Friday night or Saturday morning for a weekend run.
Last edited:
stopcrazypp
Well-Known Member
Not written out anywhere. I just copied the graph to a graphics editor, crop and resized the graph so that the pixels are proportional to the numbers and then looked at what coordinate each point on the graph is. It's the easiest way to get fairly accurate numbers from any graph.
Very true. I like djp's approach too looking at time as it shows the diminishing returns better.
sublimaze1
8Dec2012 / Leeroy Jenkins
I just dropped my slider down to about 75% today. So when I get in the car, I am at about 200 range. I figure I only drive 75-100 miles per day. And even if I decide to head over to Ft Worth on a whim, I can still make it. It is an occam's razor type of thinking, on my part. That is, will charging at 75% cause harm: will charging 92% cause harm: will I need more than 75%: can I live with less than 92%. All philosophical (non numerical) reasoning points to charging at 75 unless otherwise directed.
djp
Model 3 Performance
I see - I copied your approach to get rate values for 20,40,60,80,100% SOC at 25C and then fit a power curve to it so I could get degradation rate estimates for in between charges (85%,90%,95%, etc.).
Unless my math is wrong, you are correct in that there are diminishing returns for degradation rate as SOC changes, and also diminishing returns in terms of "what % charge is left after X years". HOWEVER, since degradation (apparently) happens as a fractional (1/2) power with respect to time, if you look at "How many years till I only have X% battery charge left", there are no diminishing returns (in fact, the opposite).
e.g., if you look at how much % capacity you have left after 8 years, dropping from 100 to 90 SOC gains you almost 4% battery capacity (75.36% to 79.13%), whereas dropping from 70% to 60% SOC only gains you 2% (84.81% to 86.81%), so there are diminishing returns in that regard. HOWEVER, if you ask "how many years till I only have 80% battery capacity left?", dropping from 100 to 90 SOC only gains you about 2 years (5.27 to 7.34 years), while dropping from 70% to 60% SOC gains you almost 5 years (13.86 to 18.4 years).
You can see this "gain in returns" in djp's numbers as well.
Obviously, the numbers themselves may be off, but if these things really behave the way we think they do in terms of exponents, etc., the general trends should be the same. For my consideration (I need to stay above, say 80% capacity), it looks like I will greatly extend the life of my car by charging to a lower percentage (charging to 90% gives me ~7 years, whereas charging to 60% gives me longer than the probable life of the car (18 years).
My spreadsheet is here: https://docs.google.com/spreadsheet/ccc?key=0Avzo7GXo-PMxdFdMcXpiV0J3OE5QNU1ITnZNSzYyaFE&usp=sharing
And now I've exhausted my "willingness to do math" quota, but I think I'll do some diligence to charge as minimally as possible. Probably will do 50% all the time unless I think there's a chance I'll want to do a long trip the next day.
Similar threads
