Tesla batteries will live longer than expected, survey finds | Hacker News Reader

Tesla batteries will live longer than expected, survey finds

engadget.com

95 pointsdoener60 comments

Tesla used to (still does?) put larger batteries in their cars than ordered, and limits them with software:

http://www.thedrive.com/sheetmetal/12820/tesla-is-silently-p...

So if these numbers are from the owners reading them from the dash... they wouldn't know their 40kwh battery was actually a 60kwh battery that was software limited. That's a lot of spare capacity to hide aging.

Remember when they flipped a switch to allow Florida owners to use the range of their larger batteries to escape the hurricane:

https://techcrunch.com/2017/09/09/tesla-flips-a-switch-to-in...

Larger batteries than purchased are clearly quite widespread with Tesla.

Then why does range falls in the first 50k miles, but it slows down after that? If that was a big part of the equation, it would be other way around: range being totally stable for first few years while 'invisible' part of battery degrades, then starting to fall off the cliff.

If you want to go full tin-foil, maybe they drop the range early to help sell the deception.

The 60kWh models with 75kWh battery packs were only sold for a brief period of time, I think their logic was that it cost less to produce a single battery pack for both the 60 and 75. They were identical with the exception of the software limit.

The capacity of all Tesla batteries is larger than the stated one primarily to hit their battery life span targets people who took them apart during salvage report about 15-20% potential higher capacity. It’s essentially the same overprovisioning we do with hard drives and SSDs.

Here is the EPA filing for the Tesla Model S 60 which shows a 65.9kw/h and a 95.5kw/h for the 85/85D. (based on the filing date of 2013..)

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=31841&...

These are dyno load tests (on the battery) not the actual road/drive tests which are used for the final classification of the car this one is used for the classification of the power train only.

As other people have noted in this discussion, the curve of battery capacity over time does not jive with an overprovisioning strategy.

Both HDDs and SDDs show no almost no degradation until they suddenly have a ton of degradation. (I've owned 3,000 SDDs, 15,000 HDDs, and 1 Model S.)

It doesn’t jolt with their EPA filings which indicate higher capacity for each model time and time again I’ve linked to one from 2013 below and it’s the same story every time.

The degradation pattern is ofc different because SSDs or HDDs are essentially binary the sector or cell is either good or bad and when you go out of spare sectors you are screwed.

Also most of them have many other modes of failure which do not involve their primary capacity.

Not saying it's impossible - or even unlikely - but has anyone actually done a tear-down confirming there are identical batteries in cars sold as having different batteries?

The reason I ask is a lot of lithium cell suppliers say "This is an x kWh battery... but don't charge above 90% or discharge below 20%" so the usable capacity is 70% of the advertised capacity. Could be that Tesla just advertises the usable capacity.

Hmm, I think it's not relevant because those constraints apply on the cell level, and Tesla battery packs are just thousands of small cells.

Well I suppose it depends if you leave x% of cells unused (prolonging their life, and ready to swap in when other cells lose capacity, much like SSDs do) or distribute across all cells, lowering the charging level X%. Either way I suspect you're hitting the "full cycle" that diminishes the cells life, so probably better to use the former strategy, leaving you with some healthy cells in the future.

Curious to hear if someone suspects otherwise though.

They don't have cell-level control, so it's definitely the latter. The cells are wired in strings, then paralleled.

http://qnovo.blogspot.co.uk/2014/11/31.html

(although apparently they do have individual fuses!)

Jason Hughes has done some battery teardowns. He actually powers his house with solar and an array of Model S battery packs. The guy's a bit nuts with the EE.

http://skie.net/skynet/projects/tesla/view_post/20_Pics+and+...

The S40 (Model S 40 KWh) was discontinued very quickly, probably no more than a couple thousand ever sold. The only other time Tesla has sold software limited batteries was when they were selling the 60 kWh models with 75 kWh software limited batteries. This only lasted for a few months between the time they started using 75 kWh batteries and when they discontinued the 60 kWh model, maybe another ~20k.

Tesla has sold around 300k Model S and Model X vehicles, so the vast majority are not software limited.

The Volt behaves in a similar way to protect against degradation, by design. I wonder if the 90% tesla s claim holds for their largest capacity cars as well.

What claim? Every Tesla suggests that you only charge to 90%, no matter how large the battery.

For some lithium battery chemistries (eg. LiIon) they have a limited shelf life regardless of number of cycles (which is why, for example, phones and tablets tend to drop sharply in battery life after 2-3 years regardless of level of usage). I'd be curious to know how the Tesla packs perform in that regard.

LiIon refers to a whole range of different chemistries; Tesla uses Lithium Nickel Cobalt Aluminum Oxide which offers a longer lifespan while typically laptops and phones use Lithium Cobalt Oxide.

I was also under the impression that Lithium batteries suffer from calendar degradation, but I wonder how much of a thing that is with modern batteries if they have good thermal management (like Teslas do), or if we even notice it when actually using a device (laptops and phones run the battery hard and hot, so end up having a 2-3 year lifespan from normal use.

Tesla use Panasonic's NCR18650B cells which were originally designed for laptops. (At least in the Model S anyway, I think the Model 3 is using something a little more custom.)

Yup, the Model 3 uses their new 2170 cells, which are definitely lithium nickel cobalt aluminum oxide [1]

[1] https://www.teslarati.com/inside-tesla-model-3-2170-lithium-...

Again, this is size not chemistry. 2170 (sometimes called 21700) refers to 21mm diameter, 70mm long. They can come in different chemistries.

Tesla does not use the standard chemistry. It is using a standard form factor.

You are describing battery sizes, not chemistries. 18650 is a standard cell size which Tesla uses in their older vehicles, and is common in older laptops and powertools. The types of chemistry used in it may vary (just like you can get AA cells in alkaline, NiCd, NiMH, and other chemistries)

No I'm not. 18650 is the size, NCR1850B is a very specific model of off-the-shelf cell in that size manufactured by Panasonic with a specific chemistry and specification. As far as I can tell (it's not something the manufacturers release much information about publicly) the "NCR" part identifies the chemistry as Lithium Nickel Cobalt Aluminum Oxide.

From what I have read on the tesla forums, batteries degrade much faster when fully charged and fully discharged. Both of these are actively discouraged by the manuals and software. For example, you have the option of setting the max charge when plugging it in which most people set to something like 80% as they normally don't need the full range.

Teslas show 0% when they're actually at 20% -- that's how full discharge is discouraged.

Overcapacity?

That's on the low side. On the high side, they apparently show the actual high side, with a warning to not go over 90% unless necessary.

Well the data shows just how much active cooling helps to lengthen their live. Gen 1 Leaf packs didn't have active cooling and their life goes downhill quickly.

https://insideevs.com/nissan-issues-statement-on-leaf-30-kwh...

No Leaf has active cooling, including the latest 40kWh Gen4.

I'm not trying to nitpick, I really cannot tell and don't have context or pre-existing knowledge to resolve it for myself and I'm not a native speaker either:

Do you mean "No Leaf has active cooling", or do you mean "No, Leaf has active cooling" (with a comma)?

Because I suspect the comma version, because I think the no-comma version would probably have been formulated slightly differently?

The reply you're responding to was a rebuff of the original parent to the thread - so it should be read without a comma i.e. it is correctly written.

You could write it as:

"Not a single Leaf has active cooling" or "None of the Nissan Leaf's have active cooling".

The above two alternative sentences aren't as elegant or concise as the original, hence why it was likely chosen.

That's surprising to me as a Ford Focus EV owner. I know that even if the supposedly inferior FFE has an actively cooled battery pack even before the 2017 refresh.

Why would Nissan fail at something so basic? Smacks of "compliance car"

I read a forum post of someone commenting on range loss while the car was parked, and others confirmed this. I thought the battery conditioning was only while the car was in use, but it seems like the battery is constantly temperature regulated even when parked. It is common in the UK to drive to the airport and leave your car parked while you go on holiday for a few weeks, but it seems that you might come back to a car that won’t get you home.

Netherlands - Belgium. Temperatures there seem to get close to freezing, but mostly not below. At least it's not California, so the batteries do hold well.

Is there a survey from places with actual winters anywhere?

Also, are these self-reported numbers? Or were actual measurements taken by those who did the study?

I believe there's been a fleet of Model S taxis in Montreal Canada for several years. They could have some cold weather data. Oh, and of course Tesla sells a ton of cars in Norway which can get chilly too.

Is this survey more than just reading the stats of the dashboard? How do we know those are correct? How much reserve do the Tesla battery packs have that is not reported in the car dashboard?

If Tesla has a reserve, the most obvious implementation would have a curve of degradation that would give it away: 100% for a long time, then a significant and steady drop.

Hard disks and SSDs work that way; hard disks have a 20% reserve these days.

1.2 Billion kilograms of lithium are estimated in reserves on planet earth - estimated by the USGS.

10 Kilograms of lithium are needed for each Tesla battery

~122M Tesla batteries can be made.

0.5M teslas produced per a year by 2020 is their aim.

20 to 100 gigafactories would mean 50 to 300 years of lithium availability.

8 year durations on Tesla batteries.

I'm not sure extended duration on these things means anything environmentally relevant.

I think your data is bit off:

https://minerals.usgs.gov/minerals/pubs/commodity/lithium/mc...

53M tons of lithium available = 53000M of kg = 5300M of Tesla batteries can be made...

Also, that's just the lithium deposits we know about now.

Actually, no, that's the lithium deposits that we know can be extracted at the current price. If the price goes up, the reserves rise immediately, because more of what we know becomes extractable.

“reserve” estimates refer to current economic viability.

The lithium production market is still a fairly new market, so even though USGS and others have made some reserve and resource estimates already, it is premature to make any firm conclusions about future resource discovery and production. We’ve learned at least one lesson from the peak oil debate: higher prices do indeed spur innovation and increased production, at least they have in the U.S. Interestingly, almost the entire increase in global oil production has come from an increase in U.S. oil production. That is, the rest of the world has barely managed to stay in place on the treadmill of oil production, even in the face of far higher oil prices.

With respect to the notion of peak lithium, there is a substantial difference between this commodity and oil because there are many ready substitutes for lithium that can be used in the manufacture of batteries. The USGS report cited by OP states: “Substitution for lithium compounds is possible in batteries, ceramics, greases, and manufactured glass. Examples are calcium and aluminum soaps as substitutes for stearates in greases; calcium, magnesium, mercury, and zinc as anode material in primary batteries; and sodic and potassic fluxes in ceramics and glass manufacture.” These substitutes may all face similar resource constraints as lithium, but there is a fundamental difference between changing battery chemistries and finding substitutes for oil more generally.

Source: https://www.greentechmedia.com/articles/read/is-there-enough...

> aluminum soaps

what kind molecule is an aluminium soap? I've tried googling but it's kinda useless, it's just soap advertisement for pages.

Dictionary says

> aluminum soap

> noun, Chemistry.

> any of the salts formed by higher carboxylic acids and aluminum, as aluminum oleate, aluminum palmitate, and aluminum stearate.

This is all completely irrelevant, and has been discussed to death. The price of the lithium in the Tesla battery is on the order of $150. If it goes up to $1500 because we run out of easily extractable resources, it's going to be economical to extract it from seawater, where there is 230 trillion kilograms of it. Also, it might be economical to recycle the lithium from old batteries (currently it's just not worth it).

I'm sure extracting lithium from sea water would have to pass serious environmental testing.

Shouldn't be much different from extracting salt from seawater, which we already do.

We already do it when making salt. We just throw most of the none NaCL salts away.

Why? It could be simply done in the countries where there are no eco-nazis, which is most of the world's countries - especially countries in warm seas where it shall be done.

I assume the "eco-nazi" part is why you're being downvoted, but I think your basic premise is absolutely correct. If the Western world outlaw this practice it will just happen in places like China and India instead.

And in a whole lot dirtier and environmentally damaging way.

With significantly less environmental oversight and regulations. The pollution could perhaps end up crossing over to western countries at higher levels than any domestic, well-regulated, production would emit.

The other side effect of added regulations is generally that the extraction gets more efficient too in order to maintain higher profits, so it helps stretch out the resource for longer too.

Lithium extraction is already done from brine, effectively from "fossil seas".

What's the recyclability of Lithium in batterie?

Probably depends on how much an organisation wants it done and who can do it: https://duckduckgo.com/?q=iphone+recycling+robot

Meanwhile, China has recently stopped accepting a lot of recyclable waste: http://www.abc.net.au/news/2017-12-10/china-ban-on-foreign-r...

Lithium’s atomic number is 3. I’m fairly certain there is a lot more of it. Because it is not a gas like hydrogen or helium, very little must escape earth.

Your numbers are off. Tesla requires 9kg of lithium per 10kWh. If 85kWh model has in reality 95kWh, then you'd need 85kg lithium per car. World reserves of lithium are roughly 16 million tons, so, it would suffices for a bit under 200 million Teslas.

Lithium production was 43000 tons last year, which would produce 500.000 cars, even if we wouldn't need lithium for anything else.

https://electrek.co/2016/11/01/breakdown-raw-materials-tesla...

https://www.statista.com/statistics/268790/countries-with-th...

https://www.statista.com/statistics/606684/world-production-...

...yet, my last two laptop batteries stopped holding a charge for more than 30 min after one year (I'm only slightly exaggerating)

Reply on news.ycombinator.com