Bigger Battery = Longer Warranty

[Joel]

This is correct, part of the decline will be due to cycling losses and another part of the decline will be due to limited calendar life. Even though these are custom-built cells, we don't know to what extent they have been optimized for EV applications. Should the cells be similar to regular laptop batteries, then assuming calendar life of six to seven years is a prudent thing to do. Calendar life itself is highly dependent on storage conditions and having a thermal management system, which can be found in all vehicles Tesla designed to date, should help maximize the useful life of the pack.

NREL did a study on vehicle preconditioning, which includes data on average capacity fade on page 6. I've copied the relevant graph below.

Thanks for the graph and information. However, this went over my head. Can you summarize this for non-technical people like myself

 

[surfingslovak]

Thanks for the graph and information. However, this went over my head. Can you summarize this for non-technical people like myself

Think of the battery as a consumable item, part of its capacity, perhaps 1-2% or so on average, will be lost every year even if it just sits. Since degradation happens due to a chemical process, the particular amount of capacity fade will be dependent on battery storage temperature. The lower the temperature, the better. Tesla vehicles have a built-in thermal system, which will keep the battery cool if you keep the vehicle plugged in. This works much like a regular refrigerator would.

The other part of the capacity fade will be dependent on usage. Each time you charge or discharge the battery, a tiny fraction of its capacity will be lost. If you go on a long trip, charge the battery to full and then discharge it to almost empty, you will put fair amount of stress on the battery. In this scenario, a larger fraction of battery capacity will be lost than when going 20 miles to the office and back. Please keep in mind that these losses are very small, fraction of a fraction of a percent per charge or discharge event.

If you expect to drive a lot of miles or plan to keep the car for a decade or more without upgrading the battery, it might be worth following a few simple rules to minimize these losses and extend the life of the pack.

Shallow cycling the battery and keeping it between 40 and 60% full will help guarantee its long life. Partial small charges and discharges will cumulatively get significantly more energy out your battery over its lifetime than deep cycles would. That's why a larger pack is advantageous. If your vehicle had 300 miles of range, you won't need to push the battery to its limits very often.

There was an excellent post last year, which included very detailed battery care recommendations by Dan Myggen from Tesla. It's worth reading, even if you don't plan to optimize the life of your battery pack.

Tesla Roadster Battery Care

 

[Joel]

Makes sense. Thanks!

 

[Kipernicus]

The lower the temperature, the better. Tesla vehicles have a built-in thermal system, which will keep the battery cool if you keep the vehicle plugged in. This works much like a regular refrigerator would.

The low temp is good only for long term sitting around, neither driving nor charging up, right?

 

[raymond]

Do 18650's last 500 full cycles even when kept at a near-optimum temperature like the liquid cooling is supposed to do?

Temperature pampering helps; DOD (depth-of-discharge) pampering probably helps more. I would be surprised to see less than 1,000 cycles from a typical Tesla battery pack.

 

[gg_got_a_tesla]

Tesla Roadster Battery Care

Fascinating stuff! Thanks! I'd want to start charging only after midnight or some such for another reason too - time-of-use metering.

 

[surfingslovak]

The low temp is good only for long term sitting around, neither driving nor charging up, right?

Batteries are usually rated for operation at room temperature. However, operating them at 40 or 50 F is not harmful in any way. Yes, lower temps will temporarily decrease pack capacity, but the difference will be fairly small - on the order of 10% for a 20 F drop. I would consider this to be good target temperature for storage as well.

Things will get a bit dicy below 15 F, and it's recommended to warm up the pack before charging or operating the vehicle in very low ambient temperatures. Since both the Roadster and the Model S have a battery temperature management system, this should be taken care of for you.

The graph I've referenced below illustrates how the available cell capacity changes with temperature. Although the battery under test was rated 6 Ah, the lowest capacity was measured at 14 F (5.5 Ah) and the highest capacity was at 140 F (7.75 Ah). This corresponds to a variation of 40% over the allowed operating temperature band

Battery calendar life can be roughly modeled using Arrhenius law. Here is what it supposedly looks like for lithium manganese cells. You can see that life expectancy is significantly better at 50 F (~ 13 years) than at 70 F (~ 8 years). These cells are similar, but not identical to what Model S will be using. The best source for this type of data wold be the manufacturer, which is Panasonic for Tesla batteries.

 

[Robert.Boston]

Sooo... here's the scenario. I drive a fairly small distance during the week, but often make long runs on the weekend. The logical pattern of charging, based on what I'm seeing here, would be for me to charge to be ready to go on the weekend, but not recharge during the week. OTOH, I live in Boston, and it can be cold overnight. If I plug in overnight, can the car (Model S) draw current to keep the battery in a comfy temperature range, but not recharge it? I'd rather drive the car during the week with SOC ~30-40%, based on what I'm reading.

 

[gg_got_a_tesla]

Sooo... here's the scenario. I drive a fairly small distance during the week, but often make long runs on the weekend. The logical pattern of charging, based on what I'm seeing here, would be for me to charge to be ready to go on the weekend, but not recharge during the week. OTOH, I live in Boston, and it can be cold overnight. If I plug in overnight, can the car (Model S) draw current to keep the battery in a comfy temperature range, but not recharge it? I'd rather drive the car during the week with SOC ~30-40%, based on what I'm reading.

A driving/usage pattern similar to mine. But, here's what I thought I understood:

1. Charge every night. Shallow cycles are better than the stressful deep cycles that your charging-for-the-weekend-from-a-relatively-depleted-pack-from-the-week's-driving suggests.
2. Schedule the charge timer such that the pack reaches the max SOC in "standard" mode (~87%) about 30 minutes before you need to leave. Those 30 min help balance the pack.
3. Even if not driving at all any given day, leave the car plugged in so that the pack is maintained at an optimal temp. (Don't know though if that means all day long on an inactive weekend day).

 

[surfingslovak]

A driving/usage pattern similar to mine. But, here's what I thought I understood:
1. Charge every night. Shallow cycles are better than the stressful deep cycles that your charging-for-the-weekend-from-a-relatively-depleted-pack-from-the-week's-driving suggests.
2. Schedule the charge timer such that the pack reaches the max SOC in "standard" mode (~87%) about 30 minutes before you need to leave. Those 30 min help balance the pack.
3. Even if not driving at all any given day, leave the car plugged in so that the pack is maintained at an optimal temp. (Don't know though if that means all day long on an inactive weekend day).

Yes, this is my understanding as well. Charging in standard mode and completing the charge shortly before driving off should be fine.

If you wanted to optimize battery life and your driving pattern allowed a lower state of charge, then you could charge to 60% or 70% via a timer. This translates to a good degree of autonomy and provides additional peace of mind, since you are doing something for the battery.

I usually charge to 70% on weekdays, and complete my commute at about 40% SOC. On weekends and holidays, I typically charge to 80% or to full, depending on our specific needs. Unless the vehicle was close to 80%, or whatever my target SOC was, I never hesitate to get an opportunity charge. I try not to dip below 30% SOC, unless I have to.

I don't know how exactly Tesla's TMS worked, but I found some information about the Volt. Hopefully it's comparable, since the Volt is using lithium ion cells with similar characteristics. Depending on the ambient temperature, the TMS could take fair amount of energy to operate. Consequently, it would be beneficial to keep the car plugged in, especially in extreme weather.

Volt thermal management system temperature band?

 

[JRP3]

If temps are moderate and your daily trips are short I would not plug in every night. There is no good reason to take the pack to a higher SOC if you don't need the range.
The cell electrolytes are voltage sensitive and are happiest in the middle ranges. Charging the cell to a higher SOC increases the time spent at higher voltages and accelerates electrolyte solvent breakdown. If you can cycle your pack between 40-60% most of the time the electrolyte stays in it's most stable voltage range. A123 has done testing with it's LiFePO4 chemistry that shows over 10,000 cycles when kept in this range. Higher temperatures also degrade the electrolyte faster than lower temperatures.
The electrolytes also breakdown overtime because they produce a compound, (which I can't remember off hand), just sitting around. As far as I can tell calender life is all related to the electrolyte, while cycle life is related to electrolyte and the cell electrodes. Cycling causes physical stresses to the electrode and also lithium plating, which takes available lithium out of solution and plates the electrodes, reducing the amount of intercalation that's possible.
Here's an interesting graph of the new Panasonic cells that Tesla will be using in the Model S. Notice the way the curve stabilizes around 80% to 70% of original capacity. This means that if you can live with around 200 miles of max range from your 300 mile pack the cells will probably last as long as the car does, discounting calender life.

 

[surfingslovak]

Here's an interesting graph of the new Panasonic cells that Tesla will be using in the Model S. Notice the way the curve stabilizes around 80% to 70% of original capacity.

Great data, thanks for posting. Note that the cells were cycled between 2.5 and 4.2V, which probably corresponds to their full capacity. They will be cycled more conservatively by Tesla's battery management system, likely between 3V and 4.1V, which should have a significant impact on battery life. Even if the pack capacity did not degrade below 70% over its life, there will likely be individual cell failures. Do you know what provisions Tesla is making for that? It would be interesting to know, how difficult or expensive will it be to replace failed cells, especially once the pack is out of warranty.

 

[JRP3]

I believe Tesla has been replacing sheets when a bad cell shows up, Roadster owners could comment further. I've been meaning to start a thread gathering data as to how many Roadster owners have had bad cells/sheets replaced.

 

[Larry Chanin]

If temps are moderate and your daily trips are short I would not plug in every night. There is no good reason to take the pack to a higher SOC if you don't need the range.
The cell electrolytes are voltage sensitive and are happiest in the middle ranges. Charging the cell to a higher SOC increases the time spent at higher voltages and accelerates electrolyte solvent breakdown. If you can cycle your pack between 40-60% most of the time the electrolyte stays in it's most stable voltage range. A123 has done testing with it's LiFePO4 chemistry that shows over 10,000 cycles when kept in this range. Higher temperatures also degrade the electrolyte faster than lower temperatures.
The electrolytes also breakdown overtime because they produce a compound, (which I can't remember off hand), just sitting around. As far as I can tell calender life is all related to the electrolyte, while cycle life is related to electrolyte and the cell electrodes. Cycling causes physical stresses to the electrode and also lithium plating, which takes available lithium out of solution and plates the electrodes, reducing the amount of intercalation that's possible.
Here's an interesting graph of the new Panasonic cells that Tesla will be using in the Model S. Notice the way the curve stabilizes around 80% to 70% of original capacity. This means that if you can live with around 200 miles of max range from your 300 mile pack the cells will probably last as long as the car does, discounting calender life.
View attachment 3305

Hi,

Thanks for the information.

I live in a hot climate where the temperature in my garage is frequently over 90 degrees. I have reserved a Model S. My daily trips are very short. I assume it would be prudent to always plug-in the car when not in use to permit battery conditioning. If the car is plugged in is it still possible to limit charging to prolong battery life? Do you have any other tips to prolong battery life in a hot climate? (Air conditioning the garage is not an option.) :biggrin:

Thanks.

Larry

 

[JRP3]

Keeping the pack cool is important, ideally you could plug the car in to allow the battery climate control to get it's power from the grid while not necessarily charging the pack if you don't need to. I don't know if the car is set up that way or not, I think storage mode might do that. One other consideration is that pack balancing needs to happen near the end of charge so occasional full charges should be done even if you don't need the range. I think Tesla actually recommends against using storage mode on a regular basis because of this in the Roadster but I'm thinking the newer Panasonic cells should need less balancing, and in truth balancing should only be an issue if taking the pack to extremes.

 

[surfingslovak]

I don't know if the car is set up that way or not, I think storage mode might do that.

Yes, based on what I've heard and read, I was thinking along the same lines. Would you know if Model S will have a sophisticated charge timer that would allow granular target SOC control? Ideally, Larry would set his SOC target to 60%, keep the vehicle plugged in and forget about any ambient temperature issues.

For what it's worth, the Leaf only allows two SOC settings: 80% or full. I believe that this roughly corresponds to standard mode and range mode in the Roadster. Is perhaps Tesla is ahead of the curve here? That would be quite intriguing.

 

[JRP3]

You can also control it with charge timing. Set charging to start late enough so the vehicle is only charged to 60% SOC or so before you plan to leave. I guess since you can do charged timing that answers the question about plugging in for climate control without charging. Once you figure out your charge timing and range needs you should be able to charge to whatever SOC you wish.

 

[stopcrazypp]

Sooo... here's the scenario. I drive a fairly small distance during the week, but often make long runs on the weekend. The logical pattern of charging, based on what I'm seeing here, would be for me to charge to be ready to go on the weekend, but not recharge during the week. OTOH, I live in Boston, and it can be cold overnight. If I plug in overnight, can the car (Model S) draw current to keep the battery in a comfy temperature range, but not recharge it? I'd rather drive the car during the week with SOC ~30-40%, based on what I'm reading.

A driving/usage pattern similar to mine. But, here's what I thought I understood:

1. Charge every night. Shallow cycles are better than the stressful deep cycles that your charging-for-the-weekend-from-a-relatively-depleted-pack-from-the-week's-driving suggests..

Yes, this is my understanding as well. Charging in standard mode and completing the charge shortly before driving off should be fine.

For those of you worried about the effects of deep cycling vs shallow cycling, I would agree with JRP3's statement that it's more important to minimize the amount of time your battery is in a high SOC and temperature. All standard battery life tests (the ones in the graphs you see of battery capacity vs cycles) are done using deep cycles, so you don't have to worry too much about deeper cycling ruining your rated cycle life (although if you want the battery to last more full cycles then shallower cycles do help).

On the other hand, high SOC and temperatures does directly effect calendar life of a cell quite dramatically. If you live in a cold climate, colder temperatures actually helps long term life (although you will see lower capacity in the short term until the BMS warms the battery back up to optimal operating temperature).

 

[vfx]

Man am I glad I don't drive one of those cars that get 40 miles on a charge and has a gasoline backup. Many owners drive 90 percent electric. That battery is toast!

 

[Joel]

Man am I glad I don't drive one of those cars that get 40 miles on a charge and has a gasoline backup. Many owners drive 90 percent electric. That battery is toast!

So, let me make sure I understand this. If someone bought a Volt or a Fisker and their daily commute is >= 30 miles, charging the battery requires deep cycles, which shortens the battery life. And the battery life of a Volt/Fisker is 20,000 miles (500 cycles x 40 miles).

I understand the Volt and Fisker have different battery technologies, but is this basically correct?