Hacker News Reader

Comment by hwillis | Hacker News Reader

Of course. Lithium is the third lightest element and nearly the most electronegative (ie it has a high voltage). It's the best battery material you could possibly pick! Not to mention, it's absurdly cheap and incredibly common. Lithium makes up <1% of the cost of a battery.

View on HN

Is it really that common? I was under the impression that it's somewhat scarce and, furthermore, a large amount of Lithium deposits reside in regions of relative political instability:

https://www.economist.com/blogs/graphicdetail/2017/06/daily-...

It's more common in the crust than lead, the second most common battery type, and uses ~100x less per kWh than lead acid batteries do.

Lithium is very poorly mapped. The last time the USGS surveyed it was in the 50s IIRC. For instance, that article is wildly inaccurate. It lists 6.9 million tonnes in the US. A single site, 45 miles to a side, in wyoming is estimated at up to 18 million tonnes of lithium[1]. Then again America may also count as a region of relative political instability.

I'll point out though that if you're gonna bring up political instability, around 20x (up to 60x in cellphones) as much cobalt is used in batteries as lithium, and virtually all of that comes from a single country: Congo. That's political instability. Lithium is a cakewalk by comparison, and fears about it are just manufactured and overblown.

[1] https://oilprice.com/Energy/Energy-General/New-Wyoming-Lithi...

I wish there were one of you for every article topic on here.

Fascinating stuff and great explanations, thanks!

From what I have read, lithium is poorly mapped because until recemt;u the global demand was low so there was no motivation to go out and look for it. Now that demand is skyrocketing, I bet a lot of companies are putting in a real search effort.

Yep, it wasn't until... 2013, I think(?), that batteries became the #1 consumer of lithium, driven entirely by Tesla. In 2009, only 21% of lithium was used for batteries, while 30% was used to in glass and ceramics. Glazes and such, as far as I understand.

[1] https://pubs.usgs.gov/circ/1371/pdf/circ1371_508.pdf

I've been trying to figure out whether Tesla has its head in the sand on the cobalt issue. Just today they had their quarterly conference call and they again played the same game of naming every element in the battery as if they were of equal concern, when quite clearly cobalt is the only thing to worry about. Anyway, I wrote this suspicious that you may have also pondered this. Even the spot price on the cobalt market stopped going up this year despite almost every major auto-manufacturer ramping up their plans. I just keep wondering what I am missing here.

There are a lot of factors. Outside the DRC cobalt is almost always a byproduct of nickel mining- Due to nickel supplies coming back under control (the price fell 80% in the past 10 years), cobalt prices haven't risen much. The main thing is that a LOT of new mines have opened/will open. Cobalt hadn't been mined in the US for over 30 years until now. We had no nickel mines until just a couple years ago. Now I think there are two cobalt mines opening and six nickel mines, in the US alone. In Canada and Australia even more have been opening. Tesla in particular has been very canny about securing contracts (Indonesia iirc? Plus several in the US... once they open). They've been publicly discussing cobalt for the better part of the decade and spurred a lot of investment in mines, and are very well prepared.

Note though that you're right: there is some potential for the short term (~3 years) price of cobalt to swing. Batteries make up a solid ~50% of global cobalt consumption. That's almost all due to China- in the US almost 70% of cobalt is used for metal-related stuff, mostly superalloys for turbines and jet engines. It's basically a race between new mines being able to ramp up production and the battery revolution taking off. Tesla isn't that worried because while cobalt is important, they can bear a short term cost increase. Tesla has repeatedly said they're under $200/kWh, and even at that price the battery would be $10,000 on a model 3. Of that the current cost of cobalt is probably ~<5%- if the price of cobalt quadruples it's still a total of <$2000. That's almost 6% of the cost- painful, but survivable. And that's an extreme situation.

Another reason they may seem blase is that cobalt is far from the biggest single cost in the battery. It's just the most potentially volatile. Only about 9% of the cathode is cobalt (in NCA), and nickel and even graphite are much bigger costs. Those two are significantly more stable- graphite can be made fully synthetically in a pinch (currently ~45% of battery graphite is synthetic[1]) and nickel consumption is already so huge (because of stainless steel) that even if every car was electric, it would hardly affect our nickel consumption. Cobalt prices can go up quite a lot before it actually affects Tesla's bottom line, but nickel and graphite will affect it immediately. In the end they all roughly balance out.

[1] more info: Synthetic graphite is made from high quality anthracite coal or refinery tar (asphalt, bitumen) that is heated in an oven (carbeurized, then graphitized) and then ground in a ball mill (then chemically treated, plus a whole bunch of proprietary voodoo). It's roughly the same cost as natural spheroidal graphite (the kind used in batteries- think best of the best of the best) and even cheaper sometimes, but they have different qualities. Natural graphite has a slightly higher capacity, while synthetic graphite has higher power capacity, so they use both in a blend. 90% of making batteries is picking the right blend for your application- there are hundreds of every ingredient to mix, tweak and balance. Competition is achingly incremental and fierce.

I agree that cobalt is a fairly small proportion of the total cost of the vehicle. Tesla should be able to weather many multiples in the increase in cobalt price. I'm more worried about the dreadful notion of a complete supply disruption (akin to 70s oil crisis). I don't think this is a common situation and is generally unlikely in most markets but with cobalt there is a confluence of factors that seems to make it more likely: the byproduct of nickel/copper effect, the political factor in DNC, the lead-time to new mines, the potential for alternative chemistries to pop up and obviate cobalt, the generally pessimistic view of EVs outside of the narrow pro-Tesla, California green circle, etc. I basically think it's possible that the EV growth curve (50+% Y/Y kind of stuff) is only believed by Tesla and a relatively few people, and that these people aren't particularly common in mining circles.

My general perspective is that Tesla should derisk the miners by putting some money upfront on a contractual basis. Maybe it's Panasonic that does this. I'm happy to see that the rumor is the price spike to 25$\lb was brought on by hedge fund speculators. I think that's a perfect example of how Wall Street greed is actually good. The high price signal helps motivate the entire supply chain.

Thanks for your lengthy answer.

> the political factor in DNC

heh

> My general perspective is that Tesla should derisk the miners by putting some money upfront on a contractual basis. Maybe it's Panasonic that does this.

I think they have a long term supply deal with Sumitomo Metal Mining[1] (Phillipines, not Indonesia, my bad), but I'm not sure if there's money up front.

It's my personal opinion that the industry is well prepared for a drastic increase in cobalt demand. I'll also point out that laptops and cell phones put a very large buffer in place for NCA and NMC chemistries. Laptops and cell phones use LCO which is a 100% cobalt oxide cathode. A high supply pressure will cause them to switch to low-cobalt chemistries like NMC and NCA, or even to completely non-cobalt chemistry. That'll free up cobalt for all batteries. LCO is becoming less popular but AFAIK it's still the majority of batteries by kWh and certainly by use of cobalt.

[1] https://evobsession.com/sumitomo-metal-mining-boosting-nca-u...

Given the direction things are going with electric cars, will there be enough to satisfy the battery demands on a global scale?

Yup, no problem. Like I said there's more of it than lead. A 100 kWh tesla will use <8 kg of lithium metal. NB you may see figures like 60-80 kg per car- that usually refers to lithium carbonate (or hydroxide), which is only 19% lithium. Anyway, 1 tonne of lithium metal can produce over 125 100 kWh electric cars.

In 2012 the world produced 4.7 million tonnes of lead metal from mines[1]. If that were lithium, it would be enough for 588 million cars. In 2012 the world made 84 million cars[2], so there's a safety margin of over 7. And that's without even really trying to mine lead. 7.2 billion tonnes of coal are mined each year, almost 2000x more than lead. We would need to mine .0071% as much lithium as we do coal.

[1] https://www.ila-lead.org/lead-facts/lead-production--statist...

[2] https://www.statista.com/statistics/262747/worldwide-automob...

Thanks for your input.

Reply on news.ycombinator.com