Every government on Earth is betting on electric vehicles. But can the planet actually mine enough lithium to make it happen?
Global EV sales hit 17.8 million units in 2024, representing roughly 20% of all new car sales worldwide. In 2025, the number climbed to an estimated 20.7 million units. China alone sold over 11 million electric cars in 2024 — almost half of all cars sold in the country.
Every one of those vehicles contains a lithium-ion battery. A typical EV with a 55–60 kWh pack requires about 40–50 kg of lithium carbonate equivalent (LCE). At roughly 0.85 kg LCE per kWh, a 75 kWh battery — increasingly the norm — needs about 64 kg of LCE.
Now multiply. At 20 million EVs per year and ~50 kg LCE each, that's roughly 1 million tonnes of LCE per year just for passenger EVs. Add electric buses, trucks, and grid-scale energy storage (the fastest-growing segment), and total lithium demand in 2025 sits at approximately 1.3–1.5 million tonnes LCE.
By 2030, analyst consensus calls for 3.7–4.6 million tonnes LCE. By 2035, 3.5–4.2 million tonnes. That's a tripling in a decade.
The question is simple: can the mining industry triple its output in ten years?
The good news is that lithium is not geologically scarce. According to the USGS 2025 Mineral Commodity Summaries, the picture looks like this:
Total proven reserves: ~30 million tonnes. Total identified resources (including deposits not yet economically viable to extract): ~115 million tonnes.
At a hypothetical sustained demand of 4 million tonnes LCE per year, 30 million tonnes of reserves represent about 7.5 years of supply, while 115 million tonnes of resources cover nearly 30 years. But these numbers are somewhat misleading — reserves expand as prices rise and technology improves. Every lithium price boom triggers exploration that discovers new deposits and reclassifies resources into reserves.
Lithium abundance is not the constraint. The Earth has plenty. The constraint is the speed at which it can be extracted, refined, and delivered to battery factories.
Here's where the math gets uncomfortable. Global lithium mine production in 2025 was estimated at 338,300 tonnes — up 20% year-on-year, which sounds impressive until you realize demand is already at 1.3–1.5 million tonnes LCE. The gap is filled by Chinese refinery output (processing imported ore) and existing stockpiles.
| Country | Output (2025e) | Change | Notes |
|---|---|---|---|
| Australia | ~113,000 t | +4% | Ngungaju mothballed |
| Chile | ~86,000 t | +5% | Brine expansion slowing |
| China | ~59,000 t | +12% | Lepidolite + brine |
| Argentina | ~25,000 t | +30% | New projects ramping |
| Brazil | ~16,000 t | +8% | Hard rock |
| Mali | ~11,300 t | New | Goulamina, Bougouni |
| Zimbabwe | ~10,000 t | +40% | Bikita + Arcadia |
| Others | ~18,000 t | — | Portugal, US, etc. |
For 2026, production is projected to grow another 15% to 389,100 tonnes. Argentina's output alone is expected to jump 42.5% to 35,700 tonnes as newly commissioned projects ramp up.
But here's the critical point: mine production is not the same as refined lithium supply. Australia mines the most lithium ore in the world, but the vast majority is shipped to China as spodumene concentrate for refining. The refining step is where the real bottleneck lives.
Mining is not software. You cannot spin up a new lithium mine on demand. According to S&P Global, the average lead time from discovery to production for mines started in 2020–2023 was nearly 18 years. In the United States specifically, the average is 29 years — the second longest in the world.
Even lithium-specific projects, which have attracted enormous investment, take 10–17 years for hard rock and 13–15 years for brine operations. Fast-tracked, well-funded projects with political support can compress this to 5–7 years. But "fast-tracked" is the exception, not the rule.
Consider the timeline for Thacker Pass in Nevada — arguably the most important lithium project in the United States:
Exploration, resource estimation, and early studies
Environmental impact statement, permitting, legal challenges from ranchers and indigenous groups
Construction begins after final court ruling
$2.26 billion DOE loan secured; construction continues
Phase I production: 40,000 tonnes/year of battery-grade lithium carbonate. This alone would be 8× current total US lithium output.
That's roughly 20 years from discovery to first production, for a project with strong government backing, massive funding, and a deposit sitting in a politically stable democracy. Most projects don't have those advantages.
Demand doubles every 3–4 years. New supply takes 10–17 years to materialize. This mismatch is the single most important fact about the lithium market.
In November 2022, lithium carbonate hit an all-time peak — above $80,000 per tonne in Chinese spot markets. By early 2024, it had crashed below $10,000. That's a ~90% decline in 15 months.
The immediate cause was simple: supply temporarily caught up with demand. Chinese refiners had overbuilt during the boom. EV sales growth in Europe stalled. Speculators fled. The market, briefly oversupplied, cratered.
The consequences, however, will echo for a decade:
By mid-2025, lithium carbonate began recovering — prices rose 56% from their January low of ~$10,800/t to ~$16,900/t by December. Goldman Sachs projects $13,250/t in 2026 and a steady climb to $17,077/t by 2028. The market is rebalancing.
But the projects that were killed or delayed during the bust? Those represent supply that was supposed to come online in 2028–2030. You can't un-cancel a mine. The geological work, the permitting, the engineering — once the clock stops, restarting adds years.
This pattern is well-known in commodity markets, and lithium is following it with textbook precision:
Fastmarkets projects the market shifting from a tiny 10,000-tonne surplus in 2025 to a 1,500-tonne deficit in 2026. More bearish analysts see structural deficits emerging by 2028–2029, with supply covering only 70–85% of demand by the early 2030s if delayed projects aren't restarted.
Even if every announced mine project were built on time (they won't be), there's a second bottleneck that most analyses underestimate: refining.
Raw lithium ore is useless for batteries. It must be refined into battery-grade lithium carbonate or lithium hydroxide — a complex, energy-intensive chemical process. And China dominates this step with extraordinary concentration:
| Metric | China's Share |
|---|---|
| Lithium refining capacity | ~67–72% |
| Cobalt refining | ~73% |
| Battery cell manufacturing | ~77% |
| Cathode production | ~85% |
| Anode production | ~92% |
China has 8% of global lithium reserves but controls ~67% of refining. By 2027, Chinese entities (domestic plus overseas operations) are forecast to control 50% of all lithium production — up from 35% just five years prior.
This means the supply chain works like this: Australian companies dig spodumene out of the ground in Western Australia. It gets shipped across the Indian Ocean to Chinese ports. Chinese refineries convert it to battery-grade chemicals. Chinese battery makers (CATL, BYD, CALB) turn it into cells. Those cells go into Chinese EVs, or occasionally get shipped back West.
For Western governments pursuing "friend-shoring" and supply chain resilience, this is a strategic vulnerability comparable to Europe's pre-2022 dependence on Russian gas. The US Inflation Reduction Act attempts to address this with escalating domestic sourcing requirements for EVs to qualify for the $7,500 tax credit:
| Year | Critical Mineral % Required |
|---|---|
| 2024 | 50% |
| 2025 | 60% |
| 2026 | 70% |
| 2027 | 80% |
| 2028 | 90% |
| 2029+ | 100% |
The intent is clear: force a non-Chinese supply chain into existence. But intent doesn't mine lithium. As of 2026, the United States produces barely 5,000 tonnes of lithium carbonate per year. Thacker Pass, if it hits its timeline, will add 40,000 tonnes by late 2027. The IRA requires 100% non-Chinese sourcing by 2029 — for a market that needs millions of tonnes. The math doesn't add up yet.
The most significant near-term disruption to lithium demand is sodium-ion battery technology, which uses sodium (effectively unlimited in abundance) instead of lithium. CATL, the world's largest battery manufacturer, has announced commercial-scale deployment of sodium-ion cells starting Q2 2026, beginning with the Aion Y Plus passenger vehicle.
BYD has developed sodium-ion batteries with up to 10,000 cycles of life. CATL's next-generation cells achieve 175 Wh/kg — enough for 500 km of driving range.
However, sodium-ion has important limitations:
The realistic impact: sodium-ion will slow the growth of lithium demand, not reverse it. If sodium-ion captures 15–20% of the entry-level EV and grid storage markets by 2030, it could reduce projected lithium demand by 400–600k tonnes LCE — meaningful but not transformative. The premium EV and long-range segments will remain lithium-dependent for the foreseeable future.
Traditional brine extraction is slow (12–24 months of evaporation) and wasteful (only ~50% lithium recovery). DLE technologies use specialized sorbents, membranes, or electrochemical processes to extract lithium from brines in hours rather than months, with 80–90% recovery rates and far lower water consumption.
DLE is real and scaling:
DLE could unlock massive new lithium sources — particularly the lithium-rich brines beneath the US (the Smackover Formation in Arkansas), geothermal brines in the Salton Sea, and oilfield brines across the Permian Basin. These could potentially make the United States a top-5 lithium producer. But "could" and "will" are separated by billions of dollars and years of engineering scale-up.
By 2030, an estimated 318 GWh of lithium-ion batteries will reach end of life, contributing roughly 1.6 million tonnes of battery waste. Currently, less than 5% of EV batteries are recycled in the US, and global lithium recovery rates are dismal.
Recycling capacity is growing rapidly — from 1.6 million tonnes/year in 2025 to a planned 3 million tonnes/year. But even optimistic projections suggest recycling can only meet 20–30% of lithium demand by 2050. The fundamental problem is arithmetic: you can't recycle batteries that haven't been built yet. The EV fleet is growing so fast that end-of-life batteries are a trickle compared to new production needs. The wave of recyclable batteries won't arrive until the 2030s, when the first generation of mass-market EVs (2018–2025) reaches retirement.
Pulling the data together, here's how supply and demand are projected to evolve. Supply figures assume all currently operating and under-construction projects hit their targets (optimistic). Demand figures use consensus IEA/analyst estimates.
Millions of tonnes LCE. Supply includes mine output + refining capacity utilization estimates.
The picture is clear: supply and demand are roughly balanced now, but a gap opens in 2026–2027 and widens dramatically toward 2030. Even with every announced project (many of which are speculative), supply covers only 70–85% of projected demand by 2030.
This means one of four things must happen:
In practice, it will be all four in varying proportions. But the first — another price spike — is the most certain. Commodity markets solve supply shortages through price.
DLE scales faster than expected, unlocking US Smackover and Salton Sea brines. Sodium-ion captures 25% of the low-end market. Argentina's lithium triangle becomes a reliable producer. Thacker Pass hits its timeline. China and the West reach tacit supply-chain accommodation. Prices stay elevated ($20–25k/t) but don't spike destructively. Result: supply meets 90%+ of demand, EVs stay on their adoption trajectory.
Several major projects are delayed 1–2 years. DLE contributes but doesn't transform. Sodium-ion takes 10–15% market share in storage and urban EVs. Another price spike to $30–40k/t in 2028–2029 triggers a second investment wave but doesn't arrive in time. Battery costs stop falling; some automakers delay EV launch timelines. Result: supply covers 80–85% of demand, EV adoption curves bend downward in regions without strong mandates.
Geopolitical disruption (China export controls, South American resource nationalism) chokes supply. DLE remains stuck in pilot phase. Recycling grows too slowly. Another bust-recovery cycle repeats, killing confidence. Lithium prices exceed $50k/t. Battery costs rise for the first time in a decade. Result: supply covers only 65–70% of demand. EV mandates are quietly pushed back. The "2030 targets" are missed everywhere outside China.
The EV transition is not in danger of failing because of lithium. The planet has more than enough lithium for hundreds of millions of electric vehicles. The danger is that it happens slower and more expensively than planned — creating years of volatility, price spikes, and political discomfort along the way.
The fundamental structural issue is that mining operates on geological timescales while demand operates on technology adoption curves. A mine takes 15 years to develop. EV sales double in 4 years. These two clocks are fundamentally misaligned, and no amount of policy can fully bridge the gap.
The most likely outcome is the "muddle through" scenario: periodic supply crunches drive prices up, which stimulates investment, which eventually brings supply online, which crashes prices, which kills investment, which creates the next shortage. This cycle — familiar to anyone in oil, copper, or rare earths — is the defining feature of commodity markets. Lithium is not special. It's just new enough that markets haven't yet learned to price it efficiently.
For investors, the implication is clear: lithium is a structural bull market interrupted by cyclical bears. The current price trough (2024–2026) is the valley before the next peak. For policymakers, the implication is that supply-chain diversification is urgent, DLE and recycling R&D deserve aggressive funding, and mandating EV timelines without securing the minerals to build them is a recipe for embarrassment.
For consumers? Electric vehicles will continue to get cheaper over time — but the decline won't be smooth. There may be a period in the late 2020s when battery costs plateau or even temporarily rise. The long-term trajectory is still toward electrification. But the road there runs through a lithium bottleneck that neither geology nor geopolitics will let us skip.
Data sources: USGS 2025 Mineral Commodity Summaries; IEA Global EV Outlook 2024 & 2025; S&P Global (mine lead times); Benchmark Mineral Intelligence; Fastmarkets; Wood Mackenzie; Mining Technology; Goldman Sachs commodity forecasts; Lithium Americas Corp (Thacker Pass filings); Rho Motion; IDTechEx (DLE market). All figures are estimates based on the most recent available data as of February 2026. Production and demand figures represent analyst consensus ranges; actual outcomes may differ.
Published: February 26, 2026 · Author: Claude (AI research, henrywhelan.com)