Update: China Copper Refining Share Climbs to ~48% (+4.3 pp); US Cu Import Reliance Hits 57%
Versus our MCS 2025 research ledger, USGS MCS 2026 (2025e) tightens copper midstream (+4.3 pp) and US copper net-import reliance (+12 pp to 57%). Gallium stays ~99%; graphite mine eases slightly; tungsten dilutes on Kazakhstan.
Loading interactive charts…
What changed since the research vintage
In our chokepoint commodities research ledger we treated USGS Mineral Commodity Summaries 2025 as a cross-mineral scoreboard: twenty mine, midstream, and export stages, China leading twelve, eight stages above 70% top-1 share, and gallium refining near 98%. That post answered the stock question — what physical inputs does the economy assume will always be available, and where is supply thin? This update answers the flow question the newest official vintage forces: what moved when MCS 2026 (version 1.3, May 2026) replaced 2024e estimates with 2025e?
The dashboard above is built as a vintage delta, not a second encyclopedia. Toggle Top-1 Δ, Prior → new, US reliance, Mine → mid, and Direction map. Filter by stage, direction, and sector. The punchline is not “everything got worse.” It is more precise: copper midstream tightened, US copper import exposure jumped, gallium stayed extreme, several mine monopolies eased slightly on paper as world output grew, and antimony’s China share was revised down even as US reliance rose after export bans.
The headline table: MCS 2025 research → MCS 2026 print
| Stage | Research top-1 | MCS 2026 top-1 | Δ (pp) | US reliance (prior → new) |
|---|---|---|---|---|
| Copper (refined) | China ~44% | China ~48.3% | +4.3 | 45% → 57% |
| Gallium (primary) | China ~98% | China ~99% | +1.0 | 100% → 100% |
| Cobalt (mine) | DRC ~74% | DRC ~74.2% | +0.2 | ~76% → ~76% |
| Rare earths (mine) | China ~69% | China ~69.2% | +0.2 | ~80% → ~80% |
| Copper (mine) | Chile ~23% | Chile ~23% | 0 | 45% → 57% |
| Natural graphite (mine) | China ~79.4% | China ~77.8% | −1.6 | 100% → 100% |
| Tungsten (mine) | China ~83% | China ~78.8% | −4.2 | >50% → >50% |
| Lithium (mine) | Australia ~37% | Australia ~31.7% | −5.3 | ~25% → >50% |
| Antimony (mine) | China ~48% | China ~36.4% | −11.6 (revision) | 86% → 91% |
Read the table as stages, not minerals. Copper mine share is flat while copper refine share climbs. Lithium pits pluralize while US lithium import reliance doubles into the >50% MCS band. Graphite China’s absolute tons rose (1.27 → 1.40 Mt) even as its share eased because Tanzania and Mozambique lifted world totals. Antimony’s large negative delta is partly a world-total revision, not a sudden Chinese exit — which is why the dashboard tags it revised rather than celebrating diversification.
Copper midstream is the vintage tightening
Open the Mine → mid panel. Copper is the family where the dumbbell widens against the economy’s “always available” assumption. Chile still digs about 23% of world mine copper — a plural pit map next to gallium or graphite. China refining moves from roughly 44% to ~48% of world refined output as Chinese refinery production prints 12.4 → 14.0 Mt while world refined copper rises only 27.6 → 29.0 Mt. Pair that with our copper mine vs refinery geography deep dive: the research post already showed China mining ~8% while refining near half; MCS 2026 says the midstream weight increased, not mean-reverted.
For the United States the same vintage is louder. MCS salient statistics put copper net import reliance at 57% of apparent consumption in 2025e, up from 45% in the research baseline year. That is a +12 percentage-point exposure move — larger than any top-1 concentration delta in the table — driven by refined import volumes and weaker domestic primary refining during smelter maintenance years. Markets that treat copper as a “diversified industrial metal” are not wrong about the mine map. They are underweighting the chemical and electrolytic gate that actually delivers cathode.
Copper also entered the Final 2025 List of Critical Minerals in November 2025. The designation does not create tons. It does reclassify the midstream story as a policy object rather than a background assumption.
Gallium and graphite: extreme, with opposite share math
Gallium remains the ledger’s hard ceiling. MCS 2026 states China accounted for about 99% of worldwide primary low-purity gallium production; Japan and Russia fill the rounding error. US net import reliance stays 100%. China lifted its US-directed gallium export ban for one year in November 2025 — a license and politics event, not a production-share event. The concentration map did not liberalize; the bilateral trade rule temporarily did.
Graphite is subtler. Switch to Top-1 Δ and sort by absolute move. China’s mine share eases about 1.6 pp to ~77.8% on our computed MCS country/world totals, even though USGS narrative language still frames China near the low-80s of world supply. The important vintage fact is dual: China produced more (+130 kt), and Africa produced much more (Tanzania roughly doubling to 75 kt; Mozambique recovering toward 60 kt). Share dilution from non-China growth is real. It is not yet a midstream story. Graphite anode processing remains a secondary IEA/research label near ~90% China — MCS does not restate that processing share — so the binding battery chokepoint is still chemistry, not only flake tons. See also natural graphite mine concentration.
Cobalt, tungsten, lithium: flat pits, diluted pits, pluralizing pits
Cobalt mine concentration is effectively flat: Congo (Kinshasa) at ~74% of world mine cobalt (230 of 310 kt). Indonesia’s rise to 44 kt (~14%) tightens the top-3 even while the top-1 barely moves. MCS narrative still names China the leading refined cobalt producer after DRC export-quota turbulence in 2025 — the classic stage flip (African pit, Asian chemistry) is unchanged.
Tungsten is the cleanest “easier on paper” mine story. China holds 67 kt in both 2024 and 2025e while world output rises to ~85 kt after Kazakhstan’s Boguty deposit starts (~2.4 kt). Top-1 share falls about 4 pp to ~79%. That is diversification at the margin, not a regime change — and APT/concentrate prices still spiked after China’s February 2025 export controls and US tariff moves.
Lithium mine share for Australia falls from ~37% to ~32% as world ex-US output jumps 222 → 290 kt (+31%). China mine output rises to 62 kt (~21%). The pit map is more plural. US net import reliance prints in the >50% band (from a lower research-era reading) — another reminder that mine pluralism ≠ US exposure relief when domestic chemical conversion and brine/clay projects are still scaling.
Who is more exposed under the new vintage
More exposed: cathode and wire-rod supply chains that assume copper refine capacity outside China scales with Chilean and Congolese mine growth; US buyers of refined copper watching reliance at 57%; compound-semiconductor and defense programs still 100% import-reliant on gallium chemistry; battery anode lines that confuse graphite mine dilution with spherical purified graphite independence; and antimony consumers facing 91% US reliance after China’s 2024–25 export restrictions even when China’s mine share revises lower in the world table.
Relatively less pressured on the mine map alone: tungsten consumers who can qualify Kazakh concentrate; lithium cell makers who can dual-source Australian spodumene and rising Argentine/Brazilian output; graphite flake buyers who can contract Tanzanian and Mozambican material — provided anode plants elsewhere actually qualify.
What would change the next update: China copper refine share falling back toward 40%; US copper reliance retreating under 50% for two MCS years; primary gallium capacity outside China clearing into merchant markets at double-digit world share; graphite anode processing shares (IEA-class) printing a durable decline; DRC cobalt mine share under 50% and refined cobalt outside China rising together.
Caveats and methodology
- Stages ≠ minerals. Counting “copper” once hides the refine tightening; counting every chemical form can double-count narratives. We publish mine and midstream rows where MCS supports both.
- Computed shares from rounded totals. Country and world figures are USGS-rounded; a 0.2 pp move is noise, a 4 pp move is signal.
- Secondary processing rows (graphite anode, lithium chemicals, rare-earth separation) are carry-forwards from the research/IEA labelsMCS 2026 does not republish those capacity shares.
- Antimony’s −11.6 pp mixes genuine Russia/world revisions with China share math; treat direction as revised, not as proof of rapid Western substitution.
- US net import reliance follows MCS definitions; a rising reliance number can coincide with a flat global top-1 share (copper mine).
- Critical-minerals list changes (copper added in 2025) are policy metadata, not production data.
- This post is a vintage delta. For the full twenty-stage scoreboard and sector filters, use the research ledger.
The shareable takeaway
Versus our MCS 2025 research print, the newest USGS MCS 2026 vintage says the economy’s quiet assumption — that specialty and bulk physical inputs keep flowing through a few jurisdictions — tightened where tonnage is largest: China refined-copper share ~44% → ~48%, and US copper net-import reliance 45% → 57%. Gallium stays near 99%. Graphite and tungsten mine shares ease at the margin as non-China supply grows. Lithium pits pluralize while US lithium reliance worsens. Midstream chemistry and licenses remain the binding gates — and copper just moved further through them.