Lithium is dug out of big open pit mines in Australia, put on a boat to China, then trucked to a plant where it is roasted at high temperature using sulphuric acid.
Henry Sanderson keeps returning to that route, and to the figure attached: China accounts for over 65% of processed lithium. His comparison is iron ore and steel. Australia produces the spodumene, China does the processing, and the relationship between the two countries is the shape of the market.
That route is the argument behind Volt Rush, his book about the raw materials underneath the clean energy transition. Many people used to assume that moving off fossil fuels would somehow not involve industrial processes, minerals or mining. Sanderson, now executive editor at Benchmark Mineral Intelligence and previously a commodities correspondent at the Financial Times, spends the book showing it will involve a great deal more of all three.
The energy that goes into the roasting does not disappear. It lands as emissions in the battery, and then in the car.
Why the roasting went east
These supply chains got to China because nobody in the West particularly wanted them. They were considered low value, energy intensive and polluting, exactly the work that was comfortable to send away. China then mastered the processing steps and got the cost down.
Graphite is the harder version of the same story. China has near absolute dominance in both forms: natural graphite, which is mined, and synthetic graphite, which starts from coal tar pitch or petroleum coke and is heated in furnaces to thousands of degrees. Sanderson's framing is that this level of control by one country goes beyond anything seen in fossil fuels.
Being ahead brings scale, and scale brings cost advantage, which is why the competitive question for Western automakers is not how to match China but how to make any dent at all. Subsidy is one answer, and the Inflation Reduction Act is the obvious instrument. But the United States had, days before the conversation, extended the timeline on graphite, giving carmakers longer to keep using the Chinese material. Sanderson reads that as a straightforward measure of how difficult shifting a supply chain actually is.
Underneath it sit questions he thinks the West has not answered. What is required in subsidy and policy to reduce reliance on China. Where to concentrate the effort. And what level of de-risking would count as comfortable, given that decarbonisation now needs to go faster, not slower.
Europe's targets start from close to nothing
Europe has put numbers on the record. By 2030, at least 10% of annual consumption extracted locally, 40% processed within the EU, 25% from recycled material, and no single third country supplying more than 65%.
Sanderson calls the targets very ambitious, mostly because the base is close to nothing. The policy exists. What he questions is the support behind it. Money in the United States and Europe has gone to battery factories and other parts of the chain, while mining and processing wait on permits. His example is a graphite mine in Sweden, still waiting on a Supreme Court decision on an appeal.
"Time is the one thing that we don't necessarily have."
The other missing piece is what surrounds a plant rather than the plant itself. China has industrial parks where the acids, the reagents, the infrastructure and the energy are all to hand, and where local government adds tax breaks. A Chinese official had that week described them as a well-stocked gym: everything needed is already in the room. In Europe and the United States, so many of the adjacent industries have gone that the basic questions have no answer. Who takes the waste product. Where the reagents come from.
Sanderson sets this inside a broader rethink in the West about market economies and the role of government, after decades in which, as he puts it, "the government was the problem, not the solution". That worked while China sat lower down the value chain doing things nobody considered high value. It stopped working when China moved up: the whole clean energy supply chain, and then high value electric vehicles with AI software and self-driving, exported. The advantages accumulated at the bottom were used to cut the cost of the things at the top. He thinks Germany's case is a sad one, given that the early 2000s push there was driven by a desire to switch to renewables and export solar panels and other clean technologies, most of which is now Chinese business.
LFP changed the mix without ending the argument
For years the line on LFP was that it was a chemistry for low-range cars in China. Sanderson credits BYD and CATL with a set of engineering improvements, particularly at pack level energy density, that took it far enough for Tesla and other Western buyers to adopt it. Without nickel and cobalt, it carries a cost advantage and avoids two supply chains entirely.
How far Western adoption goes against NMC is, in his view, still open. So is LMFP, which adds manganese to raise energy density and raises its own question about whether new manufacturing lines are needed to build it.
But he does not read any of this as the end of nickel and cobalt. Look out of the window and count the vehicles: trucks, and range-sensitive markets such as the United States, will need high-range batteries.
Chinese companies buy when the price is falling
Cobalt gives him his sharpest example. The Democratic Republic of Congo is "the Saudi Arabia of cobalt", and a few years ago China Molybdenum, a Chinese mining company partially owned by a local state-owned company, bought the Tenke Fungurume mine there from the American company Freeport-McMoRan. It is one of the best copper and cobalt mines in the world. Copper prices had fallen at the time. China Molybdenum has since overtaken Glencore to become the largest cobalt producer in the world.
Nickel followed a similar pattern, with Chinese companies putting billions into Indonesia to turn the country's ore into battery material.
Sanderson does not soften what comes attached. In cobalt, people digging by hand, children among them. In Indonesia, industrial parks running in a coal-based economy. He holds both halves at once: the material is needed now, it is better to make electric vehicles than not to, and the supply chains have to be improved. What he will not do is treat talk about sustainability as a substitute for money in the ground. Somebody has to build the plants and produce the nickel and cobalt, and on that count he gives the Chinese credit and would like to see Western companies engage.
The same volatility explains his reading of vertical integration. BYD is, he thinks, the only automaker producing its own batteries, alongside semiconductors and much else, and the two largest battery producers in the world are CATL and BYD. But carmakers generally stop before mining. Tesla has offtakes rather than mines. General Motors backing a lithium mine in the United States is one of the few exceptions.
He is not convinced everyone should attempt everything, since producing batteries as well as vehicles is hard enough on its own, and people who are good at mining might be left to mine. The argument that does persuade him is price risk. A cell producer holding a lithium asset can carry it through a year when it is uneconomic and be supplied by it when prices rise. A company with a single asset is exposed, and ends up shutting it down. With prices low and overcapacity in China, that is the live problem: how investment gets made at the bottom of the cycle, when Western companies and their investors turn cautious and Chinese ones historically have not.
He does not think globalisation is dead
There is a lot of talk that it is. Sanderson's position is that a globalised supply chain is still how costs come down, how specialisation happens and how talent around the world gets used, and that the political drive to localise will not marshal the productive power the transition needs.
What he expects instead is a small number of areas walled off for reasons such as military use, rare earths among them, supported with government money. For the rest, some structure that keeps the gains of the last twenty or thirty years while reducing the geopolitical exposure, whether through joint ventures or something like them.
Sodium-ion is where the arithmetic bites. What he has heard is that lithium needs to be at least 150,000 renminbi a tonne for sodium-ion to be economic, and that sodium-ion costs can run to double lithium-ion. At today's lithium price, the economic case is under question. His counter is that nobody can forecast a commodity price over a month or two years, and that low prices curtailing investment now is precisely what sets up the next rally.
The case for sodium, in his reading, is not really an economic one. China treats electric vehicles and clean energy as energy security: fewer oil imports, an industrial renaissance in advanced manufacturing, fewer choke points. That means wanting a suite of technologies, diversified supply chains, and raw materials mined domestically even where the economics do not support it. Europe and Russian gas, and supply chains during COVID, are what the alternative costs. Stationary storage, which does not carry the demands of a vehicle, is where sodium fits most easily.
Which leaves the question he ends on. Lithium-ion achieved remarkable cost reductions because of scale. What else can be scaled far enough to do the same.
This piece draws on the full conversation, which is available with a complete transcript on the episode page.