30 second story: Cobalt is the most controversial mineral traditionally required for lithium ion batteries. Most known cobalt reserves are located in the Democratic Republic of the Congo, which is definitely in the Congo river basin, and trying its best at the other parts of its name. Most saliently, cobalt mining is associated with poor conditions and child labour, which has led companies to seek cobalt alternatives. Lithium iron-phosphate (LFP) is an established alternative with less energy density, and many battery companies (including Tesla) are increasingly turning to LFP cathodes. Meanwhile, there is a strong push to develop cobalt free batteries using modern NMC (Nickel-Manganese-Cobalt) cathode technologies).
WILL TESLA COMMERCIALIZE A COBALT-FREE NICKEL CATHODE?
JANUARY 1ST 2030
Prior: 60%
Explanation: For my first post, I'm cheating. I posed this question to the president of a venture capitalist firm responsible for technical investments in battery start-ups. This VC firm is much more technical than most, and the president, James, has much more experience than I do with battery research, especially on the cathode side. He said 60%, so I'll use that (He gave in maybe one minute of thought, treat it as HIS prior). It also jives with my generally idea that this is a really hard problem that Tesla really wants to solve, as seen in this tweet from 2018:
It's important to note the "Nickel Cathode" aspect of the question, because Tesla is moving towards the established technology of LFP and there's no reason to think they won't make cars with LFP batteries if they don't want to. The James agrees that the probability for "cobalt-free" LFP anodes is closer to 95%.
Updates:
1. Personal experience.
I spoke to Jeff Dahn in 2019. Jeff is a researcher at Dalhousie University, one of the most respected battery researchers in the world, and also working for Tesla. He implied that Musk's tweet promising zero-cobalt batteries came soon after the Dahn group had demonstrated promising results of cobalt-free nickel cathodes in a specific context. However, battery performance relies on so many different metrics and conditions, and Jeff still felt like their cobalt-free cathodes were unworkable in a general context as of 2019. This contrasts slightly with some of the language in his Betteridge's Law style paper from the same year. My main takeaway was that he really wanted a shift towards LFP electric vehicles. This should push towards a lower estimate.
The fact that I was there makes this feel like privileged information, but it's actually probably in line with what a VC executive could know. Also, there's a bias to update too much on private information.
New value: 55%
2. The law of diminishing returns.
It's been relatively easy to moving from lithium cobalt oxide (~100% cobalt) to NMC111 (33% cobalt) and now to NMC811 (10% cobalt). There are fundamental reasons why having even a very small amount of cobalt would be beneficial to the alloying processes of nickel. This should push towards a lower estimate.
James understood this when choosing his estimate, and I don't think it's more important than he does.
New Value: 55%
3. Public Perception.
We haven't moved far from a coin toss. Consider Tesla being able to say loudly, "none of our cars use any cobalt". I think this would be a huge public relations with for them to avoid articles about exploitative mining, Musk's dad was an apartheid emerald tycoon, etc., regardless of how fair the articles are they are. This would push Tesla to make cobalt-free cathodes even if it compromised slightly on performance compared to "cobalt-lite" cathodes. I could tell that I thought this PR win was more important than James did, and I'm writing this blog, so...
New Value: 65%.
4. Cobalt Mining
There is still considerable interest to open cobalt mines in Canada and elsewhere, seemingly anticipating a strong need for cobalt in future battery markets. At first, this seemed like a good reason to revise downward. But there are in fact many battery applications aside from EVs (and especially aside from Teslas) and the market is expected to explode, so this seems predictable. No update.
New Value: 65%
5. Scientific Research
I'm slightly disappointed with the number of recent papers on cobalt-free nickel cathodes. The most relevant paper seems to be from 2020 by the Manthiram group at UT Austin. They use NMA (aluminum instead of cobalt) and find slightly lower capacity with similar retention. This is reminiscent of LFPs, but still maintaining higher capacity. It has been cited over 100 times in two years, nothing to scoff at, and it's out of a powerhouse group and already has a start-up associated with their technology. Other research papers seem either older or less relevant.
If Tesla were to adopt this technology or very similar technology with aluminum, the timeline required for deployment seems just barely long enough. They only need to have one vehicle using a similar technology, not all vehicles.
New Value: 70%.
Final prediction: 70%
Progress Checks:
Q1. Will Electra's cobalt refining plant in Ontario open by the end of 2022?
In the world where Tesla builds a cobalt-free anode: 85%
In the other world: 90%
Q2. Will Elon Musk announce the move to cobalt-free nickel anodes at an official battery day before January 1st 2025?
In the world where Tesla builds a cobalt-free anode: 70%
In the other world: 15%
I want to include these checks in my articles, but honestly, I forget how the Bayesian reasoning is supposed to update the math. I'll try to revisit these at the appropriate dates.
Photo: The Bugaboos, British Columbia
Comments