Tesla’s New Stance On Charging Habits + Elon Responds: Video

DEC 6 2018 BY DOMENICK YONEY 132

Sometimes you have to go all the way.

A Long-Range Tesla Model 3 with a mere 11,079 miles on the odometer should still enjoy its full 310-mile range. But, if you only charge it to between 50 and 70 percent for an extended period, you may see a huge drop in that figure. Luckily, this is both reversible and avoidable. The video above explains the whys and hows of this phenomena, but we’ll give you a quick summary.

The world of electric vehicles is not always super straightforward. Even people who have been living the battery-powered life for some time can discover new things. This is the case in the video above. Kim, from the popular Like Tesla YouTube channel, decided to charge her Tesla Model 3 to 100 percent full recently, only to find that it would only offer 260 miles of travel. The car was missing an entire 50 miles worth of range. Others have seen similar results after charging to only 70 percent for an extended period.

This was the first time she had tried to fill the battery pack all the way up. Because she had heard it said from experts that lithium batteries are happiest around 50 percent, she tried to keep her car’s charge in that neighborhood. While that information was technically correct, what makes sense in a lab doesn’t always completely translate to the car in your garage.

Puzzled and frustrated, she took her car to a service center where they told her the 70-percent charging routine was the source of the problem. They recommended charging to 90 percent daily, despite Tesla CEO Elon Musk recommending, in a tweet from March of 2014 (embedded below), charging to 80 percent for a particular customer’s situation.

After giving the car a “hard reset,” her Model 3 once again began showing its full range potential on the screen. They then suggested a new charging regimen to keep her car happy. Basically, they say that, besides charging to 90 percent regularly, one should also charge to a complete 100 percent at least once every three months. Also, it’s helpful if a car if left plugged in after it reaches a full charge, as it will take this opportunity to do some balancing work.

Being skeptical of this new recommendation, Kim decided to take her questions all the way to the top, asking Musk in a Tweet about the best state of charge practices. He replied (embedded below) that it’s not “worth it” to go below 80 percent and that 90 percent is still fine. He added that discharging to 5 percent or lower wasn’t an issue (for battery pack health).

Near the end of the video, Kim offers up a chart that demonstrates the effect of charge rates on longevity that should put Tesla owners minds at ease. It appears to show that owners might see only a 10 percent loss of range after several hundred thousands of miles. As with everything, your mileage may vary, but we think the 90 percent charge is fine and people should follow that recommendation, stress less, and enjoy their vehicles.

Source: YouTube

Categories: Tesla, Videos

Tags: , ,

Leave a Reply

132 Comments on "Tesla’s New Stance On Charging Habits + Elon Responds: Video"

newest oldest most voted
Mikey

I’ve noticed something like this in my Volt. If I don’t fully delete the battery for several weeks straight, my estimated range will go down slightly. Then if I go on a long trip and completely delete the battery for a hundred miles or more, I get the range back. I assumed it just couldn’t tell where the bottom of the pack was without getting there every once in awhile.

lo

lol, can t do that with pure ev

Scott Franco

What key you you hit to “delete battery” there tex?

dan

The gas button. Just hold it down until range shows zero.

Pushmi-Pullyu

delete ≠ deplete

#GrammarNazi

MoMac

How do you know that the poster didn’t just miss the ‘p’?
Most people posting in comment sections type quickly, and don’t proofread.

EVLatino

You never “fully deplete” or “fully charge”a Volt’s battery as by design you only have access to about 60% of its capacity (20% min – 80% max) so you never reach either the top or the bottom. The shorter range you see is solely based on your previous driving habits.

(⌐■_■) Trollnonymous

aaaaannnnd…….. That’s why we call it a “GOM” (Guess O Meter).
😛

ab

Personally I think there should be an option in the car’s UI to choose between “miles/km remaining” to “kWh remaining”, especially now that we are seeing mainstream cars with over 200 miles of range.

Scott Franco

Absolutely. Gas was in gallons, battery SOC is in kWh. At the very least display this along with the guesstimated miles of range left figures.

Eric Cote
Yeah, the problem is that remaining kWh isn’t a simple volumetric measurement like gasoline. They have to measure the voltage, and then estimate based on that voltage what the pack has left for kWh. Seems easy, but it’s complicated by the fact that the voltage isn’t just a function of pack capacity remaining, it will droop when power is being consumed, it will float higher when energy is being put back in, it will be affected by temperature and also how long it sits at a given level, and it’s not linear across the battery’s state of charge. Sometimes I think the KISS principle should be used a bit. The on-board charger should know exactly how many kWh it is putting into the battery, and so that should help with calibrate pack measurement. Maybe they already use this as well in their kWh estimates, I’m not sure. It’s still a guessing game since the battery is never truly empty, and part of the reason why a “hard reset” was referenced. Anyways, all these elements and more make a kWh measurement tough to be entirely accurate, though the measurement estimate certainly exists and could be reported to the driver on the… Read more »
Pushmi-Pullyu

Yes, it’s surprising that it’s so very hard to determine either the exact SoC (State of Charge) or the exact full capacity of a battery pack, even when an expert is doing the measurements.

This is something I expect to improve in the next few years; I expect either engineers will figure out how to directly measure capacity and SoC, or else EV makers will start using a different sort of battery (solid-state?) which can be more easily measured.

This limitation is the sort of thing for which engineers tend to find a solution, as a technology advances from the “early adopter” stage to mainstream use.

Bill Howland

Both the Tesla Service Center and Musk seem to have given good information here. I do around 90% (actually 87.5%) charging most times with my BOLT ev and only occasionally go to a full 100%.

The Bolt seems to get confused a bit on a 90% to 10% discharge – stating that the vehicle only has, say 8 miles left. I’ll charge it for 2 kwh (Input from the wall), and then the miles remaining will jump to 16. I’ll drive 5 miles and the indicator will STILL be at 16. THEN driving further seems to line up the actual juice left with the mileage indicator.

I’m sure there is plenty of data collected all along the way. The Tesla Roadster maintained HUGE files as to battery performance, and the CAR’S COMPUTERS would calculate an estimated battery size BACK into the file (very impressed with the car’s programmed deductive reasoning). Of course, the guess-o-meter in the ROADSTER seemed more accurate than even the BOLT, as it didn’t seem to ever get confused as it does occasionally with the BOLT as mentioned.

James

I do not know GM, but Tesla’s will change based on it current and historical driving. It could be that it think you are driving better and so your range isn’t going down. It’s a guess.

antrik

Virtually every Li-Ion battery (even those in cheap phones etc.) use a coulomb counter to estimate SoC; and voltage measurements only for safety and recalibration of the counter. That’s precisely why the indicator tends to lose track if you never fully discharge, as it can’t recalibrate itself in that case.

Cypress

I’ve never seen an ICE with a display that shows remaining gallons. It’s a just a gauge with E to F with graduated demarcations for 1/8, 1/4, 1/2, etc.

EVs I have driven may have battery percentage graph or similar visual. gauge. Why do you need to know the exact kWh remaining?

Bill Howland

GM PHEV products show gasoline usage in gallons. They are a bit obtusely displayed – my ELR only changing its display every 0.13 gallons. Internally I’m sure it keeps better records since you can drive 2 equal distances with the engine operating similiarly, and then the indicator will just jump up once – so that it must remember it used half that much the first time.

QCO

Just about very modern car has a “miles to empty” indication as part of the basic trip computer functionality.

LS

They already have that. Most EVs have an option to display percentage.

Pushmi-Pullyu

But it doesn’t display the real percentage.

https://cleantechnica.com/2018/08/26/the-secret-life-of-an-ev-battery/

Bill Howland

This guy is just guessing. GM has stated they don’t have a sliding window, but they’re pretty tight lipped about what they do reveal just the same.

Tesla statements here seem much more reasonable. Musk has said extreme top end charging could make a longevity difference, but don’t worry about discharging the battery too far. Just don’t let it get too cold for too long, which is exactly what GM also says.

Kosh

I would be happy if they showed you the actual % number you’re setting SOC at. What is that line exactly? is it 75? 80? So please just show us the number it’s set at.

Cypress

My spark EV has a display that shows the percentage and KWh used by the motor and accessories and the climate control system. Gives a little circle graph. So at the end of the drive, you can see how many kWh were used and calculate remaining.

BoltEV (was SparkEV)

Why? Balancing is less than satisfactory answer for 50 miles loss. Also curious if such large amount of loss happens with S/X. SparkEV didn’t (charged full less than 10 times in 30K miles), and BoltEV doesn’t so far.

Cypress

How do you set your spark EV to partial charge? Or did you do it manually or via a smart EVSE?

BoltEV (was SparkEV)

Manually, set the timer in my phone to 1 hour before full charge time shown by car. Even with that, most full charges were due to me being lazy.

Cypress

Too much hassle. I just plugin when I get home and charge up until it stops. I have a 2014, which has a 21kwh battery capacity that they limited to something like 18kwh useable, so it already has a built in 20% buffer. It’s why the later model years use a smaller 19kwh capacity battery but still have the same range rating. The useable is still 18kwh. Later models with the new battery have less built in buffer.

It also has a battery warranty and if it degrades beyond that before the warranty ends, then I get new battery. If it the warranty runs out and the battery degrades where I don’t have enough useable range for my needs I’ll sell it, or give it to family or friend that can use it, or see how much a new battery is in another 4 years.

Unplugged

I have to set my Focus Electric manually also. I try to keep it at about 70-80% charge, but, as BoltEV notes, like on his old SparkEV, I sometimes slip and charge to 100%.

If I was going to turn it in after the lease expiration in 2021, I wouldn’t care. But the lease buyback (residual) is only $8500. And since my lease cost is so low, my entire investment in the Focus Electric be less than $18K if I buy it. I guess we’ll see what’s offered in 2021.

Mil

“Because she had heard it said from experts that lithium batteries are happiest around 50 percent”

Completely untrue. Jeff Dahn and Elon had said to charge up to 70% to 80%. No one mentioned 50% and that would be ridiculous. Simply mentioning this inaccurate statement is all that is required for short sellers to write more FUD about EV!

Pushmi-Pullyu

You’re oversimplifying. The advice I’ve seen elsewhere (but mostly contradicted here) is to balance the daily charge/discharge around 50%. That doesn’t mean to charge to 50%; it means — for example — to charge to 70% if your daily drive will discharge the battery to 30%, or charge to 80% if your daily drive will discharge to 20%, etc. 50% was said to be the average (midway between daily charge and discharge) you should shoot for, not the level of charge you should start the day with.

Li-ion batteries don’t “like” to be charged to 100% or to be drained to 0%. It strains them, and doing that repeatedly ages them prematurely. And you’ll see advice for long-term storage for both BEVs and PHEVs that you should store with the battery around 50-60% charge. (That’s the official recommendation for the Volt; it’s in the manual. I presume that applies to other plug-in EVs which have li-ion packs, altho after reading this article I’m far less sure.) Again, that’s because the battery is least strained at that level. Even here, with the current advice, they aren’t recommending charging to 100% except very occasionally.

Ravilyn Sanders

“Li-ion batteries don’t “like” to be charged ”

Never anthropomorphize batteries. It is against their religion.

Pushmi-Pullyu

😆 Apologies for the heresy. 😉

Nix

LOL!!

Cypress

“Li-ion batteries don’t “like” to be charged to 100% or to be drained to 0%”

Which is why they build in bufffers and there is a lot of discussion of actual capacity vs useable capacity.

Eric Cote

Batteries prefer to be at their nominal voltage level. For many cells that is 3.7V and corresponds to roughly 50%, although there’s a lot of variation in voltage across temperature too. An 80% charge is decent because it at least avoids the non-linear portion of the battery’s voltage curve (where more stress to the battery exists) while also still keeping the voltage reasonably close to nominal as it droops in colder weather.

LS

Correct, except it is around 3.9V that is 50% and optimal. https://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries

Pushmi-Pullyu

I’m pretty sure that maximum voltage and optimal voltage vary by the exact cell type/chemistry. Li-ion batteries are not all created equal. I appreciate the link to Battery University, and I’ve often cited that source myself, but some of the info there is outdated.

antrik

The nominal/optimal/maximal voltages do indeed vary between different electrode chemistries; however, all the cobalt-based ones (LCO, NMC, NCA) have pretty much the same levels.

antrik

That article actually says that charging to 3.9 V gives 60 – 65% nominal capacity, not 50%. But that’s beside the point, since he was talking about nominal voltage, not charge end voltage…

antrik

Most cells actually have a nominal voltage of around 3.65 V — which usually ends up rounded up to 3.7 V or rounded down to 3.6 V…

AIUI, the nominal voltage is actually not the mid-point at 50% SoC, but rather the *average* voltage over the nominal capacity range. This allows calculating the nominal capacity in Wh from the nominal voltage multiplied by the nominal capacity in Ah…

Unplugged

An example: Up until I viewed this video, my seldom used Model 3 was kept at a constant charge of about 50-60%. If I was going on a trip (that’s about all we use it for), I would charge it up to a percent where it wouldn’t drop below 30% when we reached our destination, or returned home.

This is because I figured, based upon technical information I had read, it was best to keep the battery of an EV in the middle range. So yes, many of us were operating under the idea that mid range was best for EV batteries.

Cypress

Problem is semantics. Most EVs have some built in buffer. So when the car says 70-80%, it’s probably at 50-70% of capacity.

David Cary

Elon says the answer that will sell the most cars. 50% is the happiest place for a battery. That is pure fact.
Elon’s point is that it doesn’t matter all that much and driveability and liveability matter much more.

Calibration for accurate range management has nothing, zero, nota, to do with battery life. This article seems to ignore this fact. 50 miles of range were never lost. That is ridiculous.

LS

Very interesting. I believe from a purely battery chemistry perspective, closest to 50% is still optimal, but likely quite diminished gains keeping it lower than 80%.
I keep my Leaf close to 50% during the week when I’m not driving it, then charge up a bit to 60-70% before taking it out. Should be optimal, although I suspect the age-related degradation is more than the cycle degradation in the Leaf so there is likely almost zero benefit to this.

scott

Lithium batteries not being used should be kept between 30% and 60%. Charging to 100% and then letting cells balance often is advantageous. Just don’t let it sit at 100% long at all. Drive it right away. Regular charging should be to 90%.

Cypress

But the automakers already buffer like 10%, so when the onboard charge indicator says 100% it’s really at about 90%.

Jeff

When Tesla tells you “you’re charging your car wrong” , this is their version of Apple telling you you’re holding your phone wrong.

cfox

Really? You think HOW people charge has no effect on the charge? Have you seen that episode of House where the woman is using her inhaler incorrectly? Is that also like apple telling her she’s holding her phone wrong?

eject

Indeed. For the mass market application of EVs drivers shouldn’t have to give any consideration to charging levels. It is also disingenuous to ask the customers to only charge the car to 80 or 90% but still advertise the range at 100% of charge.
All this should be done internally and whatever states of charge can be reached by charging or discharging through driving simply have to be safe.

Pushmi-Pullyu

That’s like claiming that gasmobile drivers shouldn’t have to give any consideration to checking the oil level, shouldn’t ever need to replace the oil filter, don’t need to regularly clean/replace the air filter, nor occasionally have the spark plugs checked.

Every type of car requires some type of regular maintenance and care. In time, drivers will learn the “care and feeding” of BEVs, just as they’ve learned the “care and feeding” of gasmobiles.

eject

No car made within the last decade or two requires the driver to worry about any of this. It will pop up a message if it needs oil or a filter replacement.

Jeff

Nightly battery recharging is not considered “maintenance”. When I recharge my cell phone, I charge it until it shows 100%. When I charge rechargeable batteries, I wait until the recharger light goes green. If the battery operates at optimal range when recharged to 80%, then the car should learn to recharge up to 80% before automatically stopping, and that should be the advertised range.

Unplugged

And when you charge your cell phone to 100% every night, that’s why the battery only lasts for 2 or 3 years.

antrik

Tesla gives users the ability to use a larger portion of the battery’s capacity occasionally when needed, in exchange for putting responsibility on them not to abuse that ability. Deal with it.

James

I guess you missed the demo where if you held other contemporary phones of the time and you could also get a similar degradation, but yes, everyone changed the antenna layout to avoid it.

Model3 Owned- Niro EV TBD -Past-500e and Spark EV,

Pretty simple- drive down to 5-10% (red light/low gas tank equivalent) and charge up to 100% a few times a quarter to rebalance.
We tend to do this naturally by going to LA trips every few months.

Ron M

Seems like there should be an App on phone that monitors the health of battery and recommends when to charge and how much to charge to to get maximum performance.

Pushmi-Pullyu

Unfortunately, directly measuring the state of battery “health”, or even measuring the full capacity of a li-ion battery pack, proves to be a very difficult challenge, even for experts. (Did you know that battery capacity varies by temperature?) Hopefully this is a problem which will be solved before long, either thru some sort of sophisticated sensor system or with a change to a type of battery (solid-state?) which might be simple to directly measure.

Ron M

I still like the idea of a phone app so you can plug in the EV in the garage and set the app to charge at a particular time and to a particular level.

cesvilmal

Or part of the car’s battery management! If it already requires you to think about charging habits, it might as well recommend a charging pattern “Hey Ron, try to go to 5% and charge to 100% sometime this week”

KumarPlocher

I like that! Someone tweet that to Elon!

TJKR

This seems like an oversight on the BMS software. Tesla needs to come up with a balancing feature to calibrate predicted range and properly balance the batteries instead of relying on users to do it, it almost seems silly with the amount of tech that is in the car.

Jon

Balancing can only happen at top of charge. If you balance at any other point, the pack won’t be balanced at the top.

bradley cross

Sounds like a software problem. Yes the coding would get tricky but still only software.

Scott

It is actually a hardware issue. Cells don’t balance by limiting other cells. They balance via BMS ON or OFF once charging reaches the float stage, which won’t happen until the car is basically charged. Someone could advance the tech by limiting the charge rate of over-performing cells and hit under-performing cells (modules in Tesla’s case) but that would be a very expensive solution to a very simple fix. Just charge to full once in a while and allow the modules all the way down to the cell level to balance properly. The same was true on our old Leaf. It needed an occasional full charge to balance the cells and range noticeably increased after the fact. This will become increasingly necessary as the pack ages and cells’ performance drifts apart over time.

CCIE

The issue is that it’s very hard to tell the SOC of a lithium battery. Full and Empty are easier to detect based on cell voltage. But, anything in between is based on a proprietary formula. Errors compound in that formula’s calculation if it’s been a long time since the battery has been at one of the known states (empty or full).

TJKR

Granted but Tesla can come up with an algorithm to put the car in battary balancing mode to remind users to go to 10% below SOC before nightly charging and charge to 90-100% several times come up with top SOC level voltages for all cells. Once the software is confident on the high level, it can give you the best educated max range. To ask people to do it manually seems ridiculous if you can just come up with some software.

Robert Weekley

So, like the Chevy Volt “Engine Start” profile, to “Lubricate The Engine”, if you don’t use it much regularly!

Basical, log the charges and if none have been to 100% after so many charges, pop up a reminder to do a 100% charge. If still not done by user, do a automatic full charge after “x” more charge cycles.

BoltEV (was SparkEV)

Story shows she charged to full, which should have reset the full voltage formula. It still showed 50 miles less.

CCIE

I should have said that the measurement issue is compounded by the fact that you can’t really let the battery get to true 0% or 100% without causing damage. So, the BMS really needs to see the SOC go from as low as possible to as high as possible a few times in succession before it’s confident of the battery’s capacity and SOC.

The Volt, and likely the Bolt, exhibit the same issue if the battery only sees partial discharges over a long period. The BMS gets unsure if true SOC and plays it safe by saying it’s lower than it really is.

BoltEV (was SparkEV)

Min and max voltages on LiIon are constants and do not change as they age. “Low as possible” is determined by fixed voltage, same with “high as possible”. Only variable is if the cells are not balanced, which they should be if the car was charged to 100%.

Actually, if the car considers 100% before balance (ie, any one cell reach 100% voltage while others have not), that could result in less than full miles even if 100% is indicated. Then not letting the car sit at 100% for a long time to balance will still result low miles range. But if that’s so, that doesn’t explain this 80% or 90% or whatever to avoid reduced miles.

antrik

AIUI, balancing starts setting in as the voltage approaches the maximum level, not only once it fully reaches that level…

Pushmi-Pullyu

That’s certainly a major factor contributing to this problem, yes. It’s not so much a software problem as a physical limitation of li-ion batteries and/or a sensor problem.

Ron M

Well if there’s a solution than I think Tesla will come up with it I took a computer programming class in 1985. I’ll tell you one thing for sure I won’t be able to solve the problem.

TM3x2 Chris

There were computer classes before Al Gore invented internet???
Who knew…

Ron M

Ha Ha when I took the class I didn’t own a computer or know what a CPU was, there were students half my age that helped otherwise I probably wouldn’t have gotten a B grade. I took a total of three computer classes PL! was the first, then I took a class in Fortran and then Basic. Needless to say I never excelled in Computer Sciences.
We did learn how to send a message over the internet using DOS though no Netscape back then. Even that was difficult to do though at least for me.

Robert Weekley

Coulohm counting, is the method to track energy used vs. Spec. Counting used Ah or Wh. State of Health (SOH) is calculated by what it can presently deliver, vs what it could deliver new.

Unlike Lead Acid Batteries, which have a steeper discharge voltage drop, Lithium batteries have a high peak voltage, that drops fast for initial dran, then voltage stays more steady, until near the end, where it drops again, quickly. So simple Voltage measurements are not acurate SOC tools.

I may be wrong, but that is what my study of this has discovered.

Alaa Sadek

I like the Nightly part.

Jamarr

Does the same apply to the older model Tesla Model S. I have a 2014 Model S.

Scott Franco

Just one offhand comment: I don’t think I would ever let my M3 go below say %10 of charge. All of the negative effects of having the car discharged, besides just being a brick, scare me. Its a system, it is always using or capable of using power. For instance it cycles during the summer to cool itself just sitting in the driveway.

Pushmi-Pullyu

Occasionally (like, once every 3 months) letting the battery discharge to <10% shouldn't cause undue strain on the battery, so long as you don't leave it at that low a SoC (State of Charge). If necessary, under the circumstances described in this article, drive the battery to near-exhaustion on your way home, then plug it in immediately.

Just as with charging to 100%, you shouldn’t leave the battery pack at that SoC for long; certainly not for a full day.

Cypress

So now I gotta try and drive around awhile, close enough to my house where I’m not worries about getting back to let the car discharge to below 10%?

People don’t want to hassle with that kind of stuff. Maybe early adopter techie types will bother, but most people will just plugin and let the car charge to full each day. So the vehicle should be designed so that habit has a minimum impact on battery life and health.

Unplugged

I think you missed the suggestion. It was stated that it wasn’t too bad to let it drop to 10%, just not to leave it there. But you don’t HAVE to let it go to 10%.

Pushmi-Pullyu

The advice given here is is almost completely, utterly contrary to everything I’ve read on the subject from those who seem to be knowledgeable on the subject. And completely, utterly contrary to advice I’ve given many times on the InsideEVs forum. In particular, experts (or at least those who I thought were experts) have dismissed the idea that there might be a “memory effect” if you don’t occasionally charge to 100%, as a myth; as a holdover from NiMH or even NiCAD battery tech. Yet here the “memory effect” is officially confirmed by Tesla.

*Sigh* 🙁

I wonder, does this apply only to the Tesla Model 3, or to other Tesla cars? Or to BEVs in general?

Matthew Kennel

It might not be the memory effect in the battery cell itself, but the calibration of the monitoring system and balancing.

Pushmi-Pullyu

Yes, that’s true; this isn’t a problem with the battery cells, but with the calibration of the BMS (Battery Management System).

But the average person (and perhaps not even the average plug-in EV owner) isn’t going to understand the difference between actual battery capacity and usable capacity, or the difference between actual SoC (State of Charge) and what the car’s instrument panel displays as SoC, which will be a percentage of usable capacity, not a percentage of actual capacity.

eject

Yes exactly. Which is why I argued in my previous comment that all this stuff has to be handled internally and shouldn’t concern the driver. Your answer to that was that it naturally would have to be the drivers responsibility to care about the technical stuff.

LS

No, the memory effect you are referencing is a permanent reduction in capacity. This is not a memory effect, this is just the BMS becoming miscalibrated if it doesn’t see the top or bottom state of charge for a very long time.

Pushmi-Pullyu

Okay, okay, perhaps I shouldn’t have used the term “memory effect”, which has a specific technical meaning that doesn’t apply here. But the layman isn’t going to understand the difference, and I’ve seen the term “memory effect” used in EV forum discussions to describe exactly the problem being discussed here.

It’s like using the word “power”. Did I mean it in its technical sense, or the broader meaning it has when used in general conversation?

Perhaps we need to find a better term. Perhaps “BMS calibration imbalance”?

Jon

The real range (capacity) was not reduced, only the estimated range. This is some consequence to how they calculate capacity.

Charging to lower voltages IS good for longevity!

However, the advice to occasionally fully charge is valid; that is when cell balancing happens. With quality cells and liquid cooling, that should not be a big issue, but something done every once and a while.

Viking79

The real range was reduced because some cell modules were not charging fully. They need to balance, which can take a long time depending on balancing current (maybe days). So it is reversible by balancing the pack.

Jon

It is unlikely that balance issue results in any noticeable loss in usable capacity. That would indicate a huge balance problem! In reality, with high quality cells, liquid cooling, etc, you can go months and still be 10 or 20mV spread. To see a 50 mile range loss would mean one or more cells is many % low, that isn’t possible unless a cell is failing.

antrik

The problem is that keeping the voltages balanced (the BMS always does that AFAIK) doesn’t automatically mean the SoC are balanced…

alohart

I don’t think that Tesla is claiming that its Li-ion battery cells have a memory effect. I suspect that the charge level calculation algorithm needs to be calibrated to the low and high charge level limits occasionally. As has been correctly stated, the charge level can’t be accurately determined by the output voltage except at the low and high charge level limits where the voltage changes much faster with the charge level.

So discharging to ~10% and charging to ~90% would allow the algorithm to establish the low and high charge level limits and total usable capacity based on output voltage. Then to calculate the charge level between ~10% and ~90%, the algorithm would accumulate current flows in and out of the battery pack to calculate the charge level.

Viking79

It is not a memory effect, but cell balancing effect that I have tried to mention in the forums here for the Clarity. As cells become unbalanced the pack capacity is that of the weakest module.

scott

It’s not a memory effect, it’s various cells in the pack being unbalanced. Charging to 100% allows the BMS to balance the cells in the pack.

The only way to avoid having a BMS and having to balance the pack is to individually charge each cell. With thousands of cells in a pack this is obviously impractical.

Pushmi-Pullyu

I’ve seen it argued, apparently authoritatively (or at least I thought so), that balancing is constantly occurring as a Tesla pack is being charged. But from this article, it seems pretty clear that’s not so for the Model 3 pack, which leads me to wonder if it’s not true for Tesla’s other packs also.

I always did wonder how it was possible to individually control the charging/discharging of cells in a parallel string inside a Tesla pack, if the only connections are the parallel electrical connections and the fusible link. From this article, I guess it’s not possible.

Jon

Pushmi-pullyu,

The very nature of balancing means it can only be done at one point. Nearly every battery system, be it automotive, grid storage, or whatever, balances at top of charge.

Cells have tiny differences in capacities. If you balance them at mid charge, or bottom of charge, they WONT be balanced at top of charge. You have to pick a spot to make them even.

Doing it at top of charge maximizes energy, and prevents any cell from going over the intended voltage.

Concerning parallel cells; any cells in parallel are by default the EXACT same voltage and exactly balanced at all times!

antrik

The original hard memory effect of early NiCd cells, and the soft memory effect of NiMH and improved NiCd cells, never had anything to do with fully charging. It is what happens when a cell is recharged without having been completely drained first. Li-Ion cells do not have a memory effect, period.

Li-Ion batteries have something that can *look* a bit like a soft memory effect, if the BMS loses track of the actual capacity when the battery is never discharged to near empty. And Li-Ion batteries consisting of several cells in series can become imbalanced if they are never charged to near full. Neither of these is to be confused with a memory effect.

ffbj

Good luck Kim et al in your new endeavors. The director/cameraman, makes her look really good, which is not difficult.
I personally think that Kim has a great sense of style and taste.
Laters…

Get Real

Translation: Kim is hot?

Tom
Anyone remember exploding Dell li-ion laptop batteries? I do. Back in the day (mid 2000’s) i managed a help desk at a college that had literally thousands of school provided laptops for faculty and students. One of the first actually to go all wireless. Anyway…what a headache. The tech not so much mature at any level including wireless networking. But I digress. We went through probably a couple thousand replacement batteries per year. We heard it all, whether it be from students, faculty, the interwebs, or the vendor. Bottom line is there is no better thing than leaving it plugged it as much as possible and at full charge as much as possible. Period. Nope…don’t care what you read or heard. It isn’t even close. Any laptop treated like that typically lasted years…often past the life cycle of the laptop itself. Don’t do that and you will be replacing it somewhere between every month and every year and failure will come in as many varieties as there are pixels on an Apple IIe. Oh….after market charging cords for a laptop? Yeah…at your own risk…whether that be destroying the laptop or burning down your house or some other flavor of fail.… Read more »
Matthew Kennel

Those were probably Lithium Iron Phosphate batteries, not higher density automotive batteries.

Or maybe NiMH?

antrik

Laptops never used LFP batteries. (These didn’t even exist in mid-2000s…) Consumer electronics are generally LCO. (Possibly NCA or NMC in some newer ones? Never seen any conclusive information on that…)

Also, laptops certainly didn’t use NiMH in mid-2000s…

Jon

“Now I’m not an engineer”

No kidding. Nearly all of your statements are incorrect.

Pushmi-Pullyu

Anyone reading your entire comment will know less about the subject than he did before reading it.

I wonder if any part of that is true; certainly the majority isn’t.

antrik

Actually, the “best” policy to proliferate laptop batteries is to plug in whenever possible, and take the battery out when it’s at 70% or so… But most people are not willing to do that — thus the best *practical* policy is to plug in whenever possible, regardless of charge level. That way, the battery rarely gets discharged, and thus rarely gets recharged. Not ideal, but better than constant cycling…

Jon

I think it is important to emphasize that the actual range was not reduced, only the estimation of range.

Presumably this has something to do with how it calculates capacity.

LikeTesla seems to be implying that there should be cell balancing at lower SoC levels. If that were done, it would mean they are not balanced at top of charge. For various reasons, balancing always is at top of charge. This maximizes capacity, and prevents any cell from being over-voltage. Further, it is difficult to balance at mid charge because the SoC vs voltage slope is very flat.

The advice to fully charge to achieve balancing every so often is valid.

Shane

But I don’t like driving around with all those heavy and dangerous electrons sloshing around in my battery. If it keeps the electron pump from overheating and makes it last longer, or if keeping the battery full keeps the electrons from allowing water vapors to condense in the battery, then OK. (Sorry, I couldn’t help myself.)

Pushmi-Pullyu

What you really want to avoid is the vacuum effect from a battery that’s drained dangerously close to zero. If that vacuum gets too strong, the cells may start imploding! (And that’s every bit as true as your comment. 😉 )

MTN Ranger

I’ll have to try this. I normally charge between 75-85%. I’ve never charged up to 100% in the 7 months I’ve owned the car.

Unplugged

Me too. I avoided full charge just as much as zero charge.

hpver

Sort of reminds me of the “memory” issues some batteries had in the days before lithium ion. I wonder if the advice is the same for the Bolt and other longer range EVs? I would suspect so as the basic battery chemistry and behavior doesn’t seem that much different.

scottf200

Summary: So Prof Jeff Dahn was correct at 70% and Elon recommended 10% more (80%) for convenience. comment image

Kosh

Right, so now I’m confused. 80 or 90%?

scottf200

70% if you don’t need much for your daily and can get to your emergency locations (ie. spouse, kids, etc). Musk is saying to add 10% for convenience (limit risk, weather, etc).

antrik

His point is that even if you can get away with 70%, it’s probably not worth the risk/effort, since the difference in degradation is negligible. (Which is what I’ve been saying for a while…)

Ravilyn Sanders

I was seeing this gradual loss, from 316 miles when it was new to slowly going to 299 miles. Tire rotation to service center, and a hard reset cured it. So get a hard reset done when you rotate the tires and you are all set.

Cypress

I just plug in and let the car charge. But then again my winter range is only 65miles and I sometimes need that full amount and more.

Cypress

Does Tesla not have much built-in buffer? Why not just design with a 10-20% buffer? So the consumer doesn’t have to think about it.

David Cary

And pay for 10-20% more battery that you never get to use? 10-20% more weight also?

Jorge Aguilar

Just purchase 2015 model S 85 D, 265 miles range with 11800 miles from Tesla POV. The purchase process was very painful, almost dropped the purchase for lack of proper customer response. However the local Tesla dealer in San Antonio, especially Dalton W. Phillips safe the day assisting me and pushed the guys iin CA. Great individual.
I took the car delivery from Houston TX last Saturday with 240 miles charged and drove the car to my home in San Antonio , 227 miles way. However when was 10 miles away from home, the vehicle indicated 0 miles left, Therefore, the 240 miles indicated initially degregted whikevdriving an average of 60 mph because it did actually only 217 miles in the real world. Nonetheless, I was able to drive home safely for 10 more miles while indicating 0 miles range.
I guess, the batteries have a built in safety margins after the vehicle indicate 0 range.
I Charged the vehicle over night at home 4mph and after 12 hour, indicated 50 miles range.
Hope my first experience driving a Tesla model S help other Tesla owner with range anxiety and what it does when goes down to 0 miles range.

antrik

I don’t know how much buffer Tesla leaves intentionally when it shows 0 miles; but it’s quite possible that it simply didn’t know there was still capacity left, since it hasn’t been drained in a long time, and thus the BMS lost track. Happens with laptops etc. as well.

John Hollenberg

Very interesting graph and information, but one key caveat: this was at 20 degrees C. Living in a place (Southern California) that gets a lot hotter than that for a good portion of the year means to take the curves with a large grain of salt. For the past 15 months I have been charging my (new to me) 2013 Tesla model S85 to 50% six days a week and 70-80% on Saturday. An 80% charge still shows around 200 miles which is about right given that my one 100% charge showed 250 miles. I am going to continue going easy on the battery as it doesn’t impact my use at all. The few times I might have charged to more than 50% but forgot to do it didn’t cause a problem because I still had 125 miles and always easily made it easily to my destination and back. Your use case may vary.

antrik

Charging only to 50% is entirely pointless. It may not be much hassle, but it’s not worth even a little.

antrik

That old tweet is quoted out of context. It was a reply to a very specific either-or question — it was never meant to suggest that it’s the best policy.

TheMac19

Can an owner initiate a “Hard Reset” on their own car, without going to a service center? I have this identical problem; would love to fix it without taking the time to cycle the battery 100% -> 5% ->100% -> 5%… Many thanks.