Magnesium — material dossier

Structural facts + context. Magnesium is the single-country refining chokepoint critical mineral with the most dramatic single supply-shock event of the past decade: in September–October 2021, a Shaanxi-province electricity-rationing order at the Yulin city Pidgeon-process cluster cut Chinese magnesium-metal output ~50% and global ingot prices spiked from ~$3,000/t to ~$10,000/t inside six weeks. That single event drove magnesium onto every major- economy critical-minerals list and remains the canonical case- study for what a Chinese refining-dominance shock looks like in real time. The German Mittelstand exposure is direct via the auto die-casting base (every gearbox housing, steering wheel armature, dashboard cross-car-beam, and EV battery housing specified in AZ91 or AM50 magnesium alloy ultimately depends on Chinese Pidgeon-process ingot).

What it is

Lightest structural metal (atomic number 12; density 1.74 g/cm³, ~36% lighter than aluminium and ~78% lighter than steel; melting point 650°C). Highly reactive — burns vigorously in air — which is why it is alloyed rather than used as pure metal in structural applications.

Two commercially relevant chemistries:

• Magnesium metal (Mg, 99.8–99.95%) — primary metallic form,

cast as ingots, sows, or T-bars. Made by either silicothermic reduction of calcined dolomite ("Pidgeon process," ~85% of world capacity, dominantly Chinese) or by molten-salt electrolysis of MgCl₂ derived from brines / seawater / carnallite (Russia Solikamsk, Israel Dead Sea Periclase, US Magnesium Utah).

• Magnesium compounds (MgO, Mg(OH)₂, MgCO₃ etc.) — refractory

bricks for steelmaking ladles + furnace linings (~80% of compound demand), pharmaceutical and food-grade Mg salts, fertiliser, cement additive. Sourced from raw magnesite ore (CaO·MgO calcined to MgO) or from seawater/brine precipitation.

The IPTM register treats this as one material with two stages: mining = magnesite ore extraction (the upstream feedstock) + brine/seawater Mg recovery; refining = magnesium-metal production (the binding policy chokepoint). Compound production is not broken out separately because the steel-refractory market is served largely by Turkish + Brazilian + Russian + Slovakian producers and is not the policy-attention pole.

Traded primarily as:

• Magnesium ingot 99.9%: bulk shipments from China to global

die-casters and aluminium smelters. Pricing references China FOB Wuxi (Shanghai Metals Market quotes) plus regional premiums (~$300–600/t for delivered Europe / North America).

• AZ91 / AM50 / AM60 / AE44 alloy ingots: magnesium pre-

alloyed with aluminium, zinc, manganese, and rare-earths; shipped to die-casting plants. Premium ~$200–400/t over pure ingot.

• Magnesium powder / granules: smaller volumes for

desulphurisation in steelmaking, military pyrotechnics, and Grignard-reagent chemistry.

Where it comes from

Mine production (~32 Mt magnesite ore, 2024)

Top magnesite-mining countries (USGS MCS 2025, 2024 data):

• China: ~70% — Liaoning Haicheng cluster (~85% of Chinese

output, world's single largest magnesite reserve area). Supplies both domestic Pidgeon-process feed and global refractory market.

• Russia: ~6% — Magnezit Group (Satka, Chelyabinsk Oblast)

+ Solikamsk Magnesium Plant carnallite brines (Perm Krai).

• Korea (DPRK): ~6% — opaque reporting; large reserve

base in Tanchon area; output traded primarily through Chinese refractory channels.

• Turkey: ~4% — KÜMAŞ Manyezit (Kütahya) + Magnesit AS

(Eskişehir); strong refractory-grade exports.

• Brazil: ~4% — Magnesita Refratários (Brumado, Bahia);

RHI Magnesita subsidiary, world's largest non-Chinese refractory player.

• Slovakia: ~3% — SMZ a.s. Jelšava + Lubeník magnesite.
• Australia: ~3% — Queensland Magnesia (Kunwarara);

small but high-grade.

• Spain: ~2% — Magnesitas Navarras + Magnesitas de Rubián.

Reserves (USGS MCS 2025, contained MgO equivalent in magnesite)

• North Korea ~2.2 Bt (single largest reserve)
• China ~1.0 Bt
• Russia ~0.65 Bt
• Brazil ~0.39 Bt
• Slovakia ~0.12 Bt
• World total ~7.6 Bt

Magnesite reserves are not the binding constraint — magnesium is the eighth-most-abundant element in Earth's crust and seawater contains enough dissolved Mg²⁺ to supply current demand for millennia. The chokepoint is overwhelmingly downstream (refining cost + permitting, not ore availability).

Refining (magnesium metal, ~1.1 Mt, 2024)

Where mined ore is reduced to metallic magnesium:

• China: ~85% — Shaanxi (Yulin Pidgeon cluster, ~60% of

Chinese output) + Shanxi + Ningxia. ~250+ small Pidgeon- process plants; consolidation has been continuously underway but slowed by the September 2021 rationing aftermath.

• Russia: ~5% — Solikamsk Magnesium Plant (electrolytic,

carnallite feed) + AVISMA. Effectively unavailable to Western buyers post-Feb 2022 / sanctioned individuals on the SMP shareholder register.

• Israel: ~3% — Dead Sea Magnesium / ICL Group (electrolytic,

Dead Sea brine feed). Operational uncertainty after 2017 ICL announcement of intent to wind down — capacity has fluctuated.

• United States: ~2% — US Magnesium LLC (Rowley, Utah;

electrolytic, Great Salt Lake brine). The only US primary magnesium producer; receives Defense Production Act attention as the sole domestic source.

• Brazil: ~2% — Rima Industrial (Bocaiuva, Minas Gerais;

silicothermic, dolomite feed).

• Kazakhstan: ~2% — UKTMP (Ust-Kamenogorsk; titanium

sponge co-producer).

• Turkey: ~1% — Esan Eczacıbaşı (small Pidgeon-process,

domestic feed).

Pidgeon-process economics: silicothermic reduction is electricity- and coal-intensive — typical Chinese plant runs ~12 MWh/t Mg plus ~7 t coal-equivalent feed per tonne Mg output. This is the structural reason for Chinese dominance: cheap thermal coal + cheap dolomite + cheap labour + lax emissions standards have historically made non-Chinese silicothermic capacity uncompetitive. Western primary capacity has been declining since the 2001 closure of Pechiney Marignac (France) and the 2003 closure of Norsk Hydro Becancour (Canada).

Price state (structural, not verified this wake)

Historical reference ranges (China FOB Wuxi, 99.9% ingot, USD/t):

• Pre-2021 baseline: $1,800–2,200/t structural (CNY

~13,000–15,000/t)

• September 2021 spike: ~$3,000/t → ~$10,000/t in six

weeks; Shaanxi Yulin electricity rationing cut Pidgeon- process capacity ~50%; CNY ~70,000/t peak

• 2022 normalisation: $4,000–5,500/t through year as

capacity restarted but carryover-cost premium persisted

• 2023–2024: $2,400–3,200/t structural — gradual return

to pre-2021 baseline as Chinese capacity re-balanced

• 2025–2026: $2,300–2,800/t China FOB Wuxi reference,

Western delivered premium ~$300–600/t

Trend direction: structurally weak in 2025–2026 as Chinese Pidgeon-process supply normalised; auto die-casting demand softening with global vehicle production cycle; energy-price risk remains the dominant tail event (any repeat of 2021 Shaanxi rationing or coal-price spike would re-trigger the single-country chokepoint dynamic).

Demand structure

• Aluminium alloying: ~47% — primary use; magnesium is the

principal alloying element in 5xxx-series wrought aluminium (beverage cans, automotive sheet, marine plate) and a key alloying element in 6xxx + 7xxx series. ~5% Mg in beverage- can body sheet alone is ~15 Mt/y aluminium consumed by the US can market, contributing meaningful magnesium pull.

• Die-casting alloys: ~35% — automotive lightweighting is

the central demand-growth story. AZ91 (9% Al, 1% Zn) and AM50/AM60 (5–6% Al, low-Mn) are the workhorse magnesium die-casting alloys. End uses: instrument-panel cross-car beams (BMW, Audi standard since 1999); steering-wheel armatures (universal across European premium auto); seat frames, gearbox housings, oil pans, transfer cases, EV inverter housings, EV battery enclosures (Tesla model Y rear-floor giga-casting, Mercedes EQS).

• Steelmaking desulphurisation: ~7% — magnesium powder

injection into hot-metal ladles to remove sulphur as MgS slag; standard for clean-steel grades.

• **Reductant in titanium / uranium / zirconium production:

~4%** — Kroll process for titanium sponge consumes magnesium metal as the reducing agent; ~50% of magnesium used in this stage is recycled MgCl₂ back to electrolysis.

• Other: ~7% — anodes for cathodic protection (pipelines,

ship hulls, water heaters), specialty chemicals (Grignard reagents), pharmaceutical Mg salts, military pyrotechnics (magnesium-PTFE-Viton flares).

Demand drivers forward:

• EV gigacasting — single-piece magnesium-aluminium

structural castings (Tesla Model Y rear-floor was ~80 kg of cast aluminium-magnesium; subsequent designs by Toyota, Volvo, Ford follow the architecture). Each high-pressure giga-cast can absorb 0.5–2 kg of magnesium per vehicle.

• CAFE / Euro-7 / China VI lightweighting standards —

every gram of vehicle weight reduced via Mg substitution improves fleet fuel-economy compliance by ~0.05 g/km CO₂.

• Solid-state hydrogen storage R&D — MgH₂ is one of the

most-studied solid hydrogen carriers; pre-commercial but could become a step-change demand driver in 2030s.

• Battery anode R&D — magnesium-ion batteries (Mg²⁺

divalent intercalation) are pre-commercial; not a near-term demand factor but a long-tail upside.

Substitution risk (downward demand pressure): aluminium-only die-casting alloys can replace AZ91/AM50 in some structural applications at modestly higher mass — ~10–20% weight penalty. The 2021 spike accelerated some auto-OEM design substitution back toward aluminium, partially reversed as prices normalised 2023–2024. Carbon-fibre composites are weight-competitive but cost-prohibitive outside aerospace + premium auto.

Policy / geopolitical context

Magnesium sits at the policy-attention pole identical to graphite, tungsten, and gallium-germanium: Chinese refining dominance + clear willingness to use export controls + a demonstrated supply-shock precedent (2021 Yulin event).

Magnesium status across critical-minerals lists:

• US Geological Survey 2022 + 2023 Critical Minerals List: yes
• EU CRMA Annex II (Critical Raw Materials List, 2024): yes,

with strategic-project status for any non-Chinese primary capacity

• Japan METI critical-minerals (5 + 11 framework): yes
• Australia Critical Minerals List 2023: yes (host country

to Korab + Latrobe Magnesium pilot projects)

• UK Critical Minerals Strategy 2022 / 2023 refresh: yes
• Canada Critical Minerals Strategy 2022: yes

The supply-side stories most likely to move magnesium policy:

1. Chinese energy-policy / coal-price posture — Pidgeon- process is so coal-electricity-intensive that any Chinese carbon-intensity push (the 2021 Shaanxi rationing was ostensibly an "energy consumption double-control" measure) directly cuts world magnesium supply. Watch CCP plenum dual-control language. 2. US Defense Production Act + Inflation Reduction Act support for US Magnesium Utah — DoD has periodically funded studies of expanded Rowley capacity; no green-field capacity has yet broken ground. 3. Western greenfield primary projects — Korab Resources (NT Australia), Latrobe Magnesium (Victoria coal-fly-ash), Magonta Resources (Quebec), Western Magnesium (Nevada), Verde Magnesium (Romania), Nyrstar Auby (Belgium) pilot. Most are pre-FID; financing has been the bottleneck since 2018. 4. Russian sanctions tightening — Solikamsk + AVISMA already off-limits to most Western buyers; further restrictions would reduce non-Chinese supply ~5pp. 5. EU CRMA Strategic Project designation — the 2024 first round designated several magnesium projects (Romanian Verde Magnesium, Slovakian SMZ Jelšava expansion); EU IPCEI funding pipeline to follow.

Non-Chinese capacity in the pipeline

• Verde Magnesium (Romania, Băiţa Bihor) — proposed 90 kt/y

silicothermic plant; CRMA Strategic Project designation 2024.

• Latrobe Magnesium (Australia, Hazelwood) — pilot using

Latrobe Valley brown-coal fly-ash as feed; nameplate ~10–110 kt/y over phased build-out.

• Korab Resources (Australia, Winchester NT) — high-grade

magnesite + proposed Pidgeon plant; small (~50 kt/y target).

• US Magnesium Utah — periodic capacity-expansion studies;

Rowley flooding events (2022) interrupted continuous output.

• Western Magnesium (Nevada) — pre-FID.
• Magonta Resources (Quebec) — pre-FID; magnesite + serpentine

tailings feed.

All combined non-Chinese new capacity = perhaps 150–250 kt/y over the next 5–7 years vs current global metal output of ~1,100 kt — meaningful but not transformative; Chinese share would shift from ~85% toward ~75% if every project ships on schedule.

Concentration risks (ranked)

1. Yulin Pidgeon-cluster operational risk — the September 2021 event remains the single best calibration data point for what a Chinese magnesium supply shock looks like. Any repeat Shaanxi electricity-rationing order, coal-price spike, or environmental crackdown on coal-fired silicothermic plants would propagate within weeks to global ingot prices and within 2–3 months to die-cast component supply. 2. Chinese export-quota / licensing — magnesium has so far escaped the explicit MOFCOM export-control regime that has captured gallium, germanium, graphite, antimony, and tungsten — but it is a candidate. The October 2025 anti-dumping investigation against EU magnesium imports was the closest signal yet of escalation potential. 3. Auto-OEM design lock-in — once a vehicle architecture is qualified on AZ91 instrument-panel cross-car beams, substitution back to aluminium requires a multi-year re-engineering and crash-test recertification cycle. This raises switching costs and locks in Chinese supply dependency. 4. Russian sanctions on Solikamsk — already excluded; minor effect on global supply but eliminates one of the few non-Chinese primary sources from Western buyer lists. 5. Israel Dead Sea Magnesium operational uncertainty — ICL has signalled wind-down intent multiple times; continued operation depends on Dead Sea brine concession politics. 6. Inverse risk: aluminium die-casting substitution under sustained price spike would partially offset demand — ~10–20% of die-casting magnesium volume is technically substitutable to aluminium at ~10–20% weight penalty.

Players to know

Chinese (refining-dominant)

• Yinguang Magnesium Industry (Yulin Shaanxi; private;

China's largest single-site Pidgeon-process producer)

• Wenxi Yinguang Magnesium (Shanxi; private)
• Ningxia Hui-Ye Magnesium Industry (Ningxia; private)
• Taiyuan Tongxiang Magnesium (Shanxi; private)
• ~250+ smaller Pidgeon-process plants in Shanxi, Shaanxi,

Ningxia, Inner Mongolia — none individually large; together the world's structural magnesium-supply backbone

Non-Chinese primary producers

• US Magnesium LLC (Rowley, Utah; private; Renco Group;

the only US primary magnesium producer)

• Solikamsk Magnesium Plant / SMP (Russia; OAO; Lovozersk

+ Solikamsk; effectively offline for Western buyers since 2022)

• AVISMA / VSMPO-AVISMA (Russia; titanium-sponge co-producer

with magnesium reductant recycling loop)

• Dead Sea Magnesium / ICL Group (Israel; ICL on NYSE / TASE;

operational continuity uncertain)

• Rima Industrial (Brazil; private; Bocaiuva Minas Gerais)
• UKTMP / Ust-Kamenogorsk Titanium and Magnesium Plant

(Kazakhstan; KASE-listed; titanium-sponge co-producer)

• Esan Eczacıbaşı (Turkey; subsidiary of Eczacıbaşı Holding;

small Pidgeon-process)

Magnesite / refractory players (compound side)

• RHI Magnesita (Austria; RHIM.VI; world's largest

refractory company; integrated Brumado Brazil + Eskişehir Turkey + Hochfilzen Austria magnesite mining)

• Magnezit Group (Russia; private; Satka magnesite +

Slovakian SMZ a.s. Jelšava)

• KÜMAŞ Manyezit (Turkey; private; Yıldız Holding)
• Imerys (France; NK.PA; multi-mineral including some

magnesite)

• Calix Limited (Australia; CXL.AX; novel magnesite

calcination process)

Downstream die-casting customers (Mittelstand-relevant)

• DGS Druckguss-Systeme (Switzerland/Germany; private;

major European magnesium die-caster)

• GF Casting Solutions (Switzerland; GF.SW; Georg Fischer

AG subsidiary)

• Meridian Lightweight Technologies (Canada/Mexico; private;

largest North American Mg die-caster)

• Rheinmetall Automotive / Pierburg (Germany; RHM.DE;

housings + structural castings)

• Mahle GmbH (Germany; private; Mg-Al pistons + housings)
• ThyssenKrupp Bilstein (Germany; TKA.DE; suspension

housings)

• GP Strategies / GP Auto (US/India)

Aluminium-alloying customers

• Alcoa (US; AA on NYSE), Norsk Hydro (Norway;

NHY.OL), Rio Tinto Aluminium (UK/Australia; RIO.L), Constellium (US/EU; CSTM on NYSE), Novelis (Hindalco subsidiary; private) — all major can-sheet, auto-sheet, and aerospace-plate producers consume magnesium as their primary alloying element.

What to watch monthly

• **Shanghai Metals Market (SMM) magnesium-ingot Wuxi

spot price** (smm.cn; daily; primary global reference)

• Asian Metal magnesium weekly report (paywalled;

asianmetal.com)

• CRU magnesium quarterly outlook (paywalled)
• **China Customs HS 81041100 + 81042000 monthly export

data** (kcs.gov.cn; tracks Chinese ingot + alloy outflow)

• US Magnesium LLC Rowley operational disclosures (private

but DoD reporting via DPA-Title-III filings is occasionally public)

• **Verde Magnesium / Latrobe Magnesium / Korab Resources

FID + financing milestones** (ASX disclosures)

• EU CRMA Strategic Project status updates (ec.europa.eu;

Mg projects in 2024 first round)

• Yulin / Shaanxi electricity-rationing news (Caixin,

S&P Global Platts) — the 2021 supply-shock pattern recurrence indicator

• Auto-OEM giga-casting roadmap announcements (Tesla,

Toyota, Ford, BMW, Volvo) — demand-side leading indicator

Cross-references

• Reports: docs/minerals/reports/YYYY-MM-magnesium.md (none

yet — first monthly report should pick up after coverage_started)

• docs/minerals/materials/tungsten.md — same Chinese-refining

chokepoint pattern; tungsten is more concentrated upstream (mining + refining both ~80% China) while magnesium is the inverse profile (mining diversified at ~70% China, refining ~85% China)

• docs/minerals/materials/graphite.md — closest analogue in

the register: Chinese refining ~85% (synthetic graphite / spherical-graphite anode), single-event supply-shock precedent (Dec 2024 export controls), end-use lock-in via battery-anode design qualification

• docs/minerals/materials/silicon.md — adjacent in the

Chinese-energy-intensive-metallurgy cluster: silicon metal + magnesium metal both depend on cheap Chinese coal-fired electricity; both are exposed to "dual-control" energy- consumption rationing

• docs/minerals/FRAMEWORK.md — magnesium is the canonical

case-study for the 2x2 framework's high-concentration + high-policy-attention quadrant: durable structural demand (auto die-casting + aluminium-alloying baseline) + Chinese-refining-dominated supply + a documented supply- shock event (2021 Yulin) that proves policy-shock transmission speed.