Securing · Metric

Critical mineral midstream processing capacity

Mining ore is only the first step. Turning it into battery-grade chemicals, separated rare earths, and finished magnets is where the real control sits — and that midstream is overwhelmingly Chinese. China refines 60 to 94% of the world's battery and magnet materials, far more than it mines. Canada holds the rock (see reserves & production) but almost none of the refinery. For a country betting on critical minerals, that is a security gap, not just an industrial one.

The reserves metric showed that Canada sits on a broad critical-minerals endowment and the permitting metric showed how slowly it turns discoveries into mines. This metric is about the step after the mine gate — refining, separation, and the conversion of mineral concentrate into the battery cathodes, anodes, and permanent magnets that an electrified economy and a modern military actually consume. It is the part of the value chain that creates the most economic value and carries the most leverage, and it is the part China has spent two decades cornering. When Beijing wants to apply pressure, it does not restrict ore; it restricts processing technology and refined output — as its 2023–25 export controls on graphite, gallium, germanium, and seven rare-earth elements made plain.

China's share of rare-earth magnet production
94%
The permanent magnets in EV motors, wind turbines, and guided weapons. Up from ~50% in 2005 (IEA, 2024).
China's share of spherical graphite anode
>99%
The processed graphite inside virtually every lithium-ion battery. Less than 1% is made outside China (Benchmark, 2024).
Canada's rare-earth processing capacity
~400 t/yr
North America's only commercial REE plant (SRC, Saskatoon), online 2024 — about 0.1% of China's refined output.

Who controls the refinery

The chart below shows China's share of global processing — not mining — for the materials at the heart of the energy transition. The pattern is consistent and steep: China's grip on the chemistry is far tighter than its grip on the rock. It mines roughly 60% of the world's rare earths but separates about 90% of them and makes 94% of the magnets; it mines little of the world's lithium but refines well over half of it.

China's share of global refining & processing, by material (2024, %)
Source: IEA, Global Critical Minerals Outlook 2025 (China share of refined/processed output, 2024) for rare-earth magnets (94%), rare-earth separation/refining (~90%), graphite refining (~90%), and cobalt & lithium refining (~60% each). Refined nickel is concentrated in China and Indonesia (the latter largely Chinese-financed); battery-grade nickel sulphate is mostly Chinese but lacks a clean single-country share, so it is discussed in the text rather than charted.
Why processing, not mining, is the leverage. A mine can be opened in a friendly country in a decade; a refinery network, the metallurgical know-how, and the environmental permits behind it take far longer and cost ~50% more to build outside the incumbent (IEA). That asymmetry is exactly why export controls now target refined materials and processing technology — and why "we have the deposits" is not the same as "we are secure."

Where Canada's own ore goes

The global shares above are abstract until you follow Canada's own output out of the country. For most of the critical minerals Canada actually mines, the concentrate leaves the mine gate and the value-adding chemistry happens somewhere else — much of it in Asia, some of it inside China specifically. The exceptions, where Canada refines at home, are the legacy commodities.

Where Canadian critical-mineral output is processed
Refined in Canada Allied offshore Asia / strategic rival Averted sale (2023–24) Mine / origin Processing destination
Drag the globe to rotate it — or focus it and use the arrow keys.

The globe opens on the Arctic because that is the one vantage point holding Canada, Norway, and East Asia at once — and because these are polar routes, not trans-Pacific ones. Markers are representative facility or mining-district locations and the arcs are great-circle paths between them, not measured shipping routes; line weight is uniform because this section shows direction of flow rather than measured share (see methodology note 6). Sources as per the table below. Coastlines: Natural Earth 110m, simplified.
MineralCanada's outputWhere it is processed
Copper~515,000 t in concentrate (2024), nearly half from BCRoughly two-thirds of concentrate is exported for smelting, mainly to China, Japan, and South Korea (348,211 t exported in 2024, ~$4.0B). Domestic capacity is limited to Glencore's Horne smelter (Rouyn-Noranda) and CCR refinery (Montréal).
Nickel & cobaltSudbury basin; Voisey's Bay, NLSplit. Vale refines Voisey's Bay concentrate domestically at Long Harbour, NL. Glencore's Sudbury smelter produces nickel–cobalt matte that has shipped to its Nikkelverk refinery in Norway — an allied destination, but offshore value-add — since 1929.
LithiumSpodumene concentrate from North American Lithium (Val-d'Or, QC) — Canada's only producing lithium mineEntirely exported for chemical conversion; no lithium conversion plant operates in Canada. Offtake buyers include Piedmont Lithium (resold internationally), LG Chem (Korea), and Tesla.
Graphite~15,000 t/yr concentrate from Lac des Îles, QC — North America's only operating graphite mineSpheronization and coating into anode material is >99% a Chinese process; graphite mined anywhere is typically shipped to China to become anode. Proposed Quebec anode plants (Bécancour, Baie-Comeau) are not yet in commercial production.
Rare earthsNo producing mine (Nechalacho, NWT, paused)The pull toward China is strong even without production: in 2023 Nechalacho's owner agreed to sell its stockpiled ore to China's Shenghe Resources (also a 9.9% cornerstone investor); in June 2024 Ottawa brokered its diversion to the SRC facility in Saskatoon for $3.3M.
Uranium~24% of world production (Saskatchewan)The counter-example: Cameco refines at Blind River, ON — the world's largest uranium refinery — and converts at Port Hope, ON. A fully domestic midstream, built over decades.
Sources: Natural Resources Canada, Copper facts (2024 concentrate production and export volumes/destinations); Glencore Sudbury INO & Nikkelverk (nickel matte flows); company disclosures and press reporting for North American Lithium offtakes (2023–24), Lac des Îles capacity, and the Vital Metals / Shenghe / SRC stockpile transaction (CBC, Globe and Mail, Northern Miner, 2023–24); Cameco (Blind River, Port Hope). Potash, Canada's largest conversion success, is excluded: it is processed to finished product at the mine and has no comparable midstream stage.

The pattern mirrors the reserves picture: where the commodity is a 20th-century strength — uranium, potash, half of nickel — the midstream is at home. Where it is a battery or magnet material, Canadian rock earns mining margins here and processing margins abroad, and the finished material is bought back at import prices.

Building the alternative — from near zero

Canada operates no commercial lithium-hydroxide refinery, no battery-grade graphite anode plant, and no commercial cobalt refining for batteries; battery-grade nickel sulphate and anode material are imported. The one genuine break came in 2024, when the Saskatchewan Research Council's rare-earth facility in Saskatoon began producing neodymium-praseodymium metal — the first and only commercial rare-earth processing in North America — scaling toward roughly 400 tonnes a year. It is a real start and a real capability. It is also, against China's output, a rounding error.

Canada's first rare-earth processing capacity (SRC, NdPr metal, tonnes/year)
Source: Saskatchewan Research Council / Government of Saskatchewan (Sept 2024): commercial NdPr metal production began mid-2024 at ~10 t/month, scaling to ~40 t/month by end-2024 and ~400 t/year at full operation in 2025; ~$71M Saskatchewan + ~$30M federal funding since 2020. China's refined rare-earth output is on the order of 300,000+ t/year — the SRC plant is ≈0.1% of it.

Canada is not alone in trying to catch up. The United States is funding MP Materials' Mountain Pass refining and magnet plants; Australia's Lynas runs the largest non-Chinese separation capacity (in Malaysia and Western Australia); and the EU's Critical Raw Materials Act has fast-tracked 47 strategic projects. But the IEA's project-by-project analysis concludes that even by 2035 the top three refiners will still hold ~82% of supply — essentially back to 2020 levels. Diversification, in its words, "will not materialise through market forces alone."

Findings

Finding 1

China owns the midstream, not just the mine

Across battery and magnet materials China refines 60–94% of global supply, and its processing share consistently exceeds its mining share — ~60% of rare-earth mining but ~90% of separation and 94% of magnets. The leverage lives in the chemistry, which is far harder to relocate than a mine.

Finding 2

The graphite anode is a near-total monopoly

Over 99% of spherical/coated graphite anode — the material in essentially every lithium-ion battery — is processed in China. Even graphite mined elsewhere is typically shipped to China to be turned into anode, which is why China's December-2023 graphite export licensing rattled the whole battery industry.

Finding 3

Canada has the rock but not the refinery

Despite a broad mineral endowment, Canada runs no commercial lithium, cobalt, or battery-grade graphite processing and imports nickel sulphate and anode material. Mining strength does not equal supply-chain security — the value, and the vulnerability, sit downstream of the mine gate.

Finding 4

The build-out has started — but it is a rounding error

Canada's first rare-earth processing plant (SRC, 2024) is North America's only one and a genuine capability, but at ~400 t/yr it is ≈0.1% of China's refined output. Allied efforts (US, Australia, EU) are real yet, on the IEA's own numbers, leave the top-three refiners holding ~82% of supply through 2035.

The verdict is weak, and the gap is structural rather than geological. Canada's mineral wealth is real, but wealth in the ground is only strategic leverage once it can be refined into the materials that batteries, grids, and weapons require — and that capacity is almost entirely absent here and almost entirely concentrated in a single strategic rival. Closing it is a decade-scale industrial project requiring sustained public co-investment, offtake certainty, and allied coordination, not a market that will correct on its own.

Sources & methodology

Primary sources
Methodology notes
  1. Scope. "Midstream" means refining, separation, and conversion into battery-grade chemicals, anode/cathode material, and permanent magnets — the steps between the mine gate and the finished cell or motor. Materials covered: lithium, nickel sulphate, cobalt, graphite anode, and rare-earth separation/magnets.
  2. Processing vs mining. Shares are of processed/refined output, deliberately distinct from mining shares; the gap between them is the point. Figures are IEA 2024 estimates except spherical-anode concentration (Benchmark) and the SRC capacity (Government of Saskatchewan).
  3. Nickel. Refined nickel and battery-grade nickel sulphate are concentrated in China and Indonesia (Indonesian capacity largely Chinese-financed). The IEA does not publish a clean China-only share for nickel sulphate, so nickel is discussed in the narrative rather than charted to avoid a non-comparable bar.
  4. Comparability. The SRC ramp (chart 2) and China's refined output are different scales by design; the ≈0.1% comparison uses China's order-of-magnitude refined rare-earth output (300,000+ t/yr) against SRC's ~400 t/yr capacity.
  5. Canadian flows ("Where Canada's own ore goes"). The destination table mixes bases by necessity: copper uses NRCan trade statistics; nickel/cobalt, lithium, and graphite use company disclosures and offtake announcements rather than a comprehensive trade series; rare earths describe a single transaction. It is illustrative of direction of flow, not a measured share of each commodity's output. The globe above the table encodes exactly this and no more: every arc is drawn at the same weight, so it shows where material goes and never implies how much. Volume-weighted arcs were considered and rejected — only copper has a comparable trade series behind it.
  6. Companion workbook. The editable processing-share and SRC-capacity tables live in Securing_Midstream-Processing_Data.xlsx.
Page last reviewed July 2026 · Data current to 2024 — the latest IEA Global Critical Minerals Outlook