The price increases have been because of supply and demand mismatches and speculation that happened because of Covid as production stabilises so will the prices. This is temporary aberration which should avg out over a decade.

Well, the covid/speculation shockwaves have affected many other goods, but I don't know any other asset that got to x15 in that timeframe. (maybe there are some others, I haven't tracked all of them tbh)

So I wonder if you can really rule out that the increase in demand for EV isn't the main factor in lithium's x15 price increase. And if it's the case, I don't see how battery prices could go down in the next decade, since increasing mining isn't spontaneous and demand is not going to decrease.

Solid state, Na based, grid scale non-Li batteries, improved (i.e. cheaper) recycling of battery metals.

There are a lot of reasons that would reduce the cost of batteries even if the mining+processing of Li didn't grow (which it will).

I'm definitely pro-EV. As soon as it's practical for us I'll get one. However, I do worry that there are so many dimensions to optimising EV batteries (degradation, fast charge / discharge, stability, temperature tolerance, weight, cubic volume and price) that we might not be able to find anything that really changes the game. By which I mean a doubling of capacity, so 200kwh easily in a car, by my maths that's what we need to keep a light van (3.5t say) busy for a whole shift. Or in my case to tow a caravan 300 miles.

The only thing preventing manufacturers from putting 200KWh in a full size van or truck is cost, and with battery cost reducing 90% per decade, that should not remain a barrier.

The cost drops of the last decade will likely not continue going forward. It is more likely that lithium ion batteries will end up kind of like Nickel-based rechargable batteries: a huge spurt in cost reduction as they become mainstream, then not much after that. Lithium ion chemistries have a few more cost reductions to go, but not another 90%.

Semiconductors spoiled us into thinking we could ride one technology idea down an exponential curve forever, but most technologies don't have that kind of progression. By the way, semiconductors didn't have that kind of progression either - chips from 10 years ago look very different than today's.

It's all about the experience curve. For every doubling in production, cost drops 10-25%. There are very few exceptions to this.

Nickel battery production has decreased as they're being replaced with Lithium batteries so they're no longer riding the experience curve.

Semiconductors have a relatively constant price per chip but a regularly decreasing price per transistor.

We need about 1000TWh for the clean energy transition, half for cars and half for the electrical. There might be better substitutes for the electrical grid, but cars are going to use something that looks a lot like a lithium-ion battery, like sodium ion.

So I figure we've got solid demand from our current 500GWh/year up until at least 32TWh/year or so, 6 doublings, which should lower the price by about 90%.

Look at the other side. Tesla's battery day announcement in 2021 promises 56% cost reductions in 4 years via 6 enhancements: silicon anodes, cobalt-free cathodes, dry electrode process, structural batteries, tabless electrodes, larger cells.

Only the dry electrode process is particularly innovative or risky, but the others still add considerable cost reductions. Not all of Tesla's innovations may pan out, but they're just one of many companies and Universities researching batteries, many of whom are more innovative than Tesla is.

I'm fully on board with you with Wright's law, but that's what we have some idea of demand for, but not what we can supply. Battery supply has a lot of constraints, and they will eventually lift, but not necessarily quickly.

Also, Tesla has a very spotty track record on predicting its own innovations (yes, even on battery design), so I wouldn't take that prediction too seriously.

It wasn’t a 15x price wing, but large commodity price swings aren’t that unusual. Oil briefly went negative due to COVID that’s larger than an infinite price swing.

https://fred.stlouisfed.org/series/DCOILWTICO/

https://www.cnn.com/2021/04/20/investing/oil-prices-negative...

Not sure why you're being downvoted, but this is the right answer

You can't predict the markets.

It is difficult to predict markets fast, reliable and precisely enough to reliably beat them. It is certainly possible to predict markets and be good at it, people do it all the time

They have a lot of variability. It's the difference between predicting the weather and predicting the average weather (that is, the climate).

It's easy to predict the markets, just as it's easy to predict the weather.

I predict it will rain sometime in the next month at my address. I also predict that it will freeze hard in the same time period.

This is a useful prediction, given that I will be winterizing my yard today.

Making a prediction better than an expert's is hard. Useful predictions are often easy.

If my father in law decides to plant 1000 acres in wheat next year, he is making a prediction about the markets. Given the situation in Ukraine and with global climate, wheat prices are one of the most difficult to predict precisely. But he is a low cost producer compared to many, so he knows he is likely to make a profit whether the price is high or low. A precise prediction would be nice, but producers everywhere operate with imprecise but useful predictions all the time.

«Situation» in Ukraine is at mid-WWII level (1942), so plant the wheat. It will be tough to produce and export wheat in Ukraine for next two years at least.

If the war in Ukraine continues, oilseeds are the most profitable planting choice.

Lithium is a very small input to batteries, by mass. I forget the exact number, but I think only a few percent by weight, and it's not a major part of the cost. There's also this breakdown:

https://www.visualcapitalist.com/breaking-down-the-cost-of-a...

I would expect battery prices to stall on their drop, or even rise a bit, in the next year or two, before dropping again massively in price. There has been a lot of supply chain disruption, and an absolutely massive increase in demand. But lots more production capacity is being built.

Small input by mass but already 50% of the battery cost, from your link So if it does x15 in price, then battery goes x7.5 ? That’s still terrible for battery price

No, the cathode is 50%, and the cathode is not just lithium.

How much of the price of the battery is the raw lithium, and how much is the cost of manufacturing? For most technological goods the prices of raw inputs are almost irrelevant; you don't see chip prices going up because of the cost of raw silicon.

(Exception: solar panels, because those have the least processing over raw silicon and are therefore more price sensitive)

With the numbers on that chart, the relative molar masses of lithium and lithium carbide, and assuming 160g metallic lithium per kWh, at current exchange rates, I got $132 per kWh.

Which is a surprise, but so is a x15 increase in a raw material cost.

Two things are happening here. One, we're bringing on much more Lithium production all around the world. The price increase is coordinating that, but people can also see the writing on the wall. There's plenty of innovation happening in Lithium mining and extraction.

The other thing is we're innovating in battery chemistry. It would not surprise me if Lithium were only one of many options in ten years, depending on the storage capacity and number of cycles needed per kg. Sodium is a likely candidate, and it's as cheap and available as salt. Iron, aluminum, and others are a possibility, for stationary batteries where weight doesn't matter.

I know people here are very skeptical of every battery advance press release, and rightfully so. But nonetheless, some of that is going to pan out, and the world of energy storage is going to fundamentally change the economics for cars, airplanes, and grid storage within the next decade.

Is it likely that current cars would be compatible with batteries that are that different to not even be Lithium in 10 years?

If I look at say my laptop or phone, they definitely aren't compatible with newer batteries even if it's technically feasible so why would we expect cars to be?

Yes, it is likely. As long as the form factor is well defined or standardized, new battery chemistries are self-contained. Think of, say, how rechargeable AA batteries can fit into standard alkaline AA slots.

The reason it's not true for laptops is that batteries aren't standard form factor sizes in laptops.

> new battery chemistries are self-contained. Think of, say, how rechargeable AA batteries can fit into standard alkaline AA slots.

You might notice that Li-Ion AA batteries CAN'T be charged by Ni-Cd or Ni-MH AA battery chargers. Even a smaller change like Ni-MH to Ni-Zn requires different equipment.

It’s a software update to support different chemistries. This is how Tesla supports various pack types (Li, LFP, etc) across vehicles and stationary storage products.

To clarify: did you mean HW update? For instance, I can't imagine a software update changing my battery chemistry from Lithium Ion to Nickel Cadmium.

I think they meant that _supporting_ this type of battery is a software update. Software can't do matter-energy conversion yet.

Your car doesn't have to know about the "chemistry" of your battery. It just needs to know specifics about voltages and power draw and temperatures. This is what drivers do.

This is like suggesting that my computer can't go from AMD to nvidia because they would need to know deep specifics about AMD's proprietary chips to support them. But that's what drivers are for!

That’s because your video card sits on a standard well defined port. Where as charging an EV right now comes a bunch of plugs, voltages, charging times, etc. until a unified open standard is reached where you can go from super charging a Tesla to charging a sodium hyndai is reached, this whole ev dance will remain anti consumer.

It's "just" different charge profile; wouldn't need any hardware change.

Most chemistries can be charged using the same "constant current up to this voltage, constant voltage to the max voltage" method, just need to set the right values.

I'd bet a decent amount of money that most electric cars in the developed world will still use the Goodenough chemistry in ten years.

The theoretical energy density of sodium is too low for the weight, this isn't a matter of developing the technology.

What we can look forward to is sodium batteries becoming cheap and reliable, replacing every application of lead-acid, and most, perhaps all, stationary lithium. Great choice for house or grid-level storage, where the lower density is basically harmless.

Batteries are such a major component of an electric car that there's a sort of Rocket Equation which applies: a sodium car with the same range as a lithium one would have to carry so much more battery that one would have to add substantially more just to move the extra weight around.

This will never pencil out.

I did say ten years, rather than more.

150Wh/kg is completely fine. A small vehicle with a 100kg battery can do 95% of what is needed for the average person (or as a second family car), it's just a matter of whether the trend of legislating sane vehicles out of existence continues.

> Batteries are such a major component of an electric car that there's a sort of Rocket Equation which applies: a sodium car with the same range as a lithium one would have to carry so much more battery that one would have to add substantially more just to move the extra weight around.

This is absolutely not clear. Active lithium is just ~60g per kWh in a battery, or ~1% of the mass of a state of the art LFP battery. Even though sodium weighs 3.5 times more for the same charge, if it's use would allow just a few percent reduction in total weigh, either by reduction of inactive cathode material or some other source, it could easily beat lithium batteries.

Type of material decides on battery's energy density. It's just impossible for battery of "inferior" chemistry to get to same level of capacity.

There is a reason lead ones are not "just" heavy but also physically much bigger than same capacity lithium one.

What will (probably) happen if we get competitive sodium batteries is that they will phase out lithium from the stationary storage so there won't be that huge demand for lithium for everything. Or possibly they will become "cheap EV's battery"

Stationary storage will more like be taken over by iron batteries which are much cheaper. Sodium batteries will happen in cars because not everyone needs really long range.

Sodium is both heavier and less energetic, enough so that the equation points to a worse car: heavier than a car which goes substantially further.

Sodium cannot improve lithium by replacing it in batteries, for physical chemistry reasons which can't be negotiated.

I don't consider it likely that lithium will become so much more expensive that people will want their second-largest purchase, a car, to be inferior in the way that it must be with sodium batteries.

Rather more likely that cheap and good sodium batteries relieve pressure on lithium supply for the enormous amount of battery storage we need to handle intermittent renewables.

If a significant proportion of a lithium battery was lithium, this would absolutely be true.

By weight, something around 1% of a lithium battery is lithium.

Sodium is seen as an interesting alternative, because there are some indications that it can make that 99% that makes up the rest of the battery lighter. Even though sodium is ~3.5 times worse as a charge carrier, if it's use allows cutting enough weight from that rest, it can be a winner overall.

For example, aluminium is, for complex reasons, not compatible with every part of a li-ion battery, and so they use copper connectors internally, typically containing by weight at least 10 times as much copper as they contain lithium. Sodium batteries can use aluminum everywhere, which saves on not just cost, but also more in weight than the weight difference between equal molar amounts of sodium and lithium.

That all makes sense about the weight, and it will likely matter for airplanes. But it doesn't matter enough for cars.

LFP (lithium iron phosphate) batteries are already used by Tesla and have a gravimetric energy density of 125 Wh/kg, much lower than Lithium-ion.

CATL currently produces a sodium-ion battery at 160 Wh/kg, and they expect the next generation to reach 200 Wh/kg. Even if that doesn't come to fruition, it's already higher density than what's being used in electric cars!

Now consider that sodium-ion is less flammable and doesn't have the risk of thermal runaway. You might be able to drop weight in hazard protection systems. Battery cycle times are competitive with Lithium chemistry. Contrary to your prediction that it doesn't pencil out, I expect sodium-ion to completely displace Lithium batteries for use in electric vehicles because it pencils out so much better. It will just be so much cheaper and doesn't bottleneck on any materials as you scale up.

As someone who occasionally walks and often cycles, I really hope batteries that are even heavier than lithium don't become commonplace.

Why? compare the weight of SUVs or trucks. It's more than the battery weight, so better hope for smaller vehicles.

I hope for smaller vehicles as well. It's hard to separate "big" and "heavy" statistically because they are highly correlated, but in general, heavier vehicles means higher likelihood of fatality in a crash 1. I suppose we will be doing a natural experiment to tease out the difference over the years as heavy EVs become more commonplace and pedestrian fatalities continue to increase.

> It’s also about weight. Trucks today averaging around 5200 pounds -- that’s a jump of more than a half ton since 2000. And things get heavier as electric vehicles look to become our new normal.

> On average electric vehicles are 10 percent heavier than gas powered cars. For instance, the new electric Hummer weighs in at a staggering 9,100 pounds or more than 4 and a half tons. All that weight can have deadly consequences. Researchers have found that for every 1,000 pounds a car puts on, a pedestrian is nearly 50 percent more likely to die if they’re hit and the more a truck weighs the longer it will take to stop.

1. https://www.clickondetroit.com/news/2022/08/08/size-does-mat...

I dunno that "lighter cars" is a realistic thing to hope for at this point. I do hope to see better bumper crush zone designs, more protected bike lanes, blinking crosswalks, better in-vehicle collision detection systems, etc.

Even if a "lighter" car hits you, it's gonna be a bad day. Better to avoid the collision altogether, no? :)

I "hope" for lighter cars in the same way I "hope" we start taking necessary drastic measures to combat climate change, which is by knowing that it won't happen and advocating for change anyway.

On the other hand, I think maybe hoping that vehicles become no heavier than modern EVs is at least somewhat realistic.

>LFP (lithium iron phosphate) batteries are already used by Tesla and have a gravimetric energy density of 125 Wh/kg, much lower than Lithium-ion.

Tesla uses iron phosphate in its cheaper models, but the high end is still reliant on NMC. I'd be interested to know the maximum range available on LFP.

BYD sells a lithium-iron-phosphate battery with an energy density of 150 Wh/kg. CATL's LFP is 125.

Also, a lithium-iron-phosphate battery is a lithium-ion battery, in that the electrolyte works by lithium ionic conduction. The appropriate contrast would be "lithium-nickel".

>CATL currently produces a sodium-ion battery at 160 Wh/kg, and they expect the next generation to reach 200 Wh/kg.

This is reported as the "single-cell" energy density, and is not directly comparable to the battery pack density.

>Now consider that sodium-ion is less flammable and doesn't have the risk of thermal runaway.

Surely this isn't because of the stability of sodium, which anyone who watched that chemistry demonstration in high school knows, is significantly more reactive than lithium. I expect the risks to be similar; furthermore, the concept of a dense energy storage that has no risk of sudden release is physically unlikely.

> Tesla uses iron phosphate in its cheaper models, but the high end is still reliant on NMC. I'd be interested to know the maximum range available on LFP.

In theory 490 km, in practice, 380 km. Because they don't degrade as much, after a few years they are closer to what you call the high end models.

> Also, a lithium-iron-phosphate battery is a lithium-ion battery, in that the electrolyte works by lithium ionic conduction. The appropriate contrast would be "lithium-nickel".

Correct, I want to point out that John Goodenough was responsible for the basic research which lead to all lithium-ion batteries.

I explicitly intended LiFePO to be included, and I hope for a variety of reasons that it's much more popular than cobalt-heavy formulas in ten years.

Only now do I realise that "Goodenough" is a real person's name, and neither a placeholder name nor a case of Emphasis Via Capital Letters:

https://en.wikipedia.org/wiki/John_B._Goodenough

I have been interested in the technology from TexPower:

https://texpowerev.com/

They offer a lithium-nickel chemistry without cobalt. If it can be proven in practice, it would allow comparable density to the best car batteries in a scalable chemistry. The founder is a researcher at the University of Texas who wrote several papers on the chemistry:

https://onlinelibrary.wiley.com/doi/pdf/10.1002/adma.2020027...

https://www.sciencedirect.com/science/article/pii/S240582972...

https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/aenm.2...

> Sodium is a likely candidate, and it's as cheap and available as salt.

If you split salt into sodium and chlorine, and use the sodium for batteries, what do you do with the chlorine?

Dump it into Californians tap water supply apparently :)

On serious note, Chlorine is widely use chemical, used in cleaning products, bleaching paper and cloth, pesticides , rubber, and some solvents.

Chlorine has lots of industrial uses, so I imagine you sell it. If you produce enough as a by product the cost will go insanely low.

If you had to dump it, you'd probably react it with something cheap to make it inert. This would generate energy, it's akin to combustion with oxygen. You could use that energy to decrease the energy requirements of splitting the NaCl in the first step.

For example, you could use it to turn calcium carbonate AKA limestone into calcium chloride, which is rather harmless. You would get carbon dioxide out of this.

That doesn't sound like a good idea, we already have too much CO2.

Something that might work on a small scale could become an ecological threat on the scale of a billion electric cars.

Bleach. Clorox sells lots of it.

Use carbonate, which doesn't have this problem.

This would be preferred anyway: HCl is nasty, and carbon dioxide is not.

The world has plenty of lithium. Just takes a couple years of mining to catch up.

There’s also a limitless source in the ocean. If prices got high enough we could just filter ocean water.

At what concentration though? At what price does lithium extraction from Sea water become economically viable? It feels to me that this is an entropy problem.

AFAICT, Lithium extraction from sea water is economical at current prices. What's preventing its mass adoption is the certainty the prices will stay at current levels.

There's also sodium, which makes for almost as good batteries for much cheaper.

Not for use in motor vehicles. Very good for infrastructure though.

No, they are great for cars. For infrastructure iron batteries are much better. Iron batteries are not suitable for cars though.

> Isn't lithium mineral the main cost in building these batteries?

No, and I have no idea why anyone would think that. Raw material costs are generally not that significant, most costs in manufacturing come from labor and capital.

The lithium in a $15000 75kWh battery pack costs around $1600 at current prices. There is a lot of room for the price to fall even with rising raw material costs, as manufacturing gets more efficient and capacity is built out.

And also, rising raw material prices will cause production to expand, which will eventually push prices back down.

This source tells a different story: https://www.visualcapitalist.com/breaking-down-the-cost-of-a...

IIUC 50% of the cost of a battery is coming from lithium

Manufacturing is only about bout 25%

That's not what the infograph tells you. It says that 50% of the battery cost comes from cathode materials, which is true for some chemistries, but even in those cases that's not because of lithium, but because of cobalt.

Which is precisely why the industry is moving to chemistries that don't use cobalt, pushing the materials price back down.

Also, lithium isn't the cathode, it's the anode. The cathode is primarily nickel (III) oxide, which is unstable and so stabilized with cobalt, aluminum and/or manganese.

Technically most Li ion batteries are created in the discharged state, with the lithium in the battery provided by the cathode at the time of construction, and then moved over to the LiC6 anode via charging.

Electric car production would have ground to a halt over the last 2 years in the face of a 15x increase in raw material costs if it accounted for 50% right? That doesn't seem to have happened.

First that wasn’t a 15x price swing, you read the chart incorrectly click on the 5 year view again. That said, Price shocks happened in both directions from COVID it’s just another supply disruption not any kind of a long term trend.

As to cost, it depends on chemistry but you need ~0.16kg of Lithium metal/kWh or 16kg for 100kWh battery pack. At current prices that’s ~1,300$ worth of lithium meaning the majority of battery pack costs are still from manufacturing and other materials.

https://www.dailymetalprice.com/metalpricecharts.php?c=li&u=...

November 2020: 40k$ Today: 562k$ Okay, 14x, not 15x

But yeah, on average if you look to previous price, maybe it's more of x8 price increase. 18 month ago was the all time low of the data on tradingview, I give you that.

About the weight details, are you sure your figures are correct? First result on google gives me way higher numbers (minimum 80g/kwh, but likely more than 160g/kwh) https://www.linkedin.com/pulse/how-much-lithium-li-ion-vehic... Also, another comment in this thread gives 60g/kwh, and another 160g/kwh

> minimum 80g/kwh, but likely more than 160g/kwh

Units… 160g/kwh is 0.16kg/kWh, 80g/kWh is 0.08kg/kWh

But yes, it varies based on chemistry. 0.16kg/kWh is on the high end.

My bad, I read 16g/kwh. Nevermind.