1. Industry overview — Chinese aluminum under the 45-million-tonne policy ceiling

2025 was an extraordinarily distinctive year for the Chinese aluminum industry. On one hand, the 45-million-tonne electrolytic-aluminum capacity ceiling set by the National Development and Reform Commission (NDRC) in 2017, after eight years of buffer, finally approached its actual binding threshold — according to the China Nonferrous Metals Industry Association (CNIA), operating electrolytic-aluminum capacity reached 44.20 million tonnes at year-end, leaving incremental headroom of less than 2 million tonnes. On the other hand, global aluminum prices entered the strongest cyclical recovery since 2011, driven by the triple stack of AI-compute demand, grid upgrades, and the new-energy-vehicle (NEV) wave. The London Metal Exchange (LME) 3-month aluminum contract averaged USD 2,690 per tonne in 2025, up 11.2 percent year-on-year; the Shanghai Futures Exchange (SHFE) front-month aluminum contract averaged RMB 19,600 per tonne in 2025 and briefly broke through RMB 21,800 in the first quarter of 2026 — the highest reading since 2011.

China's electrolytic-aluminum output for 2025 was 43.70 million tonnes, accounting for 59.8 percent of global production and remaining the world's largest producer for 22 consecutive years. Alumina output reached 87.60 million tonnes, bauxite imports 158 million tonnes, and recycled aluminum 11.50 million tonnes. Globally, the International Aluminium Institute (IAI) reported 2025 primary aluminum output at 72.50 million tonnes; China alone accounted for nearly 60 percent. The remaining 29 million tonnes of ex-China primary aluminum was supplied mainly by Rusal, Rio Tinto, Alcoa, Emirates Global Aluminium (EGA), Hindalco, Vedanta Aluminium, and Norsk Hydro.

In supply-chain terms, China has assembled the world's most complete aluminum value chain — from bauxite mining, alumina refining, electrolytic-aluminum smelting, to deep aluminum processing. But three Damocles' swords still hang overhead: external dependency on raw materials, the carbon intensity of the power mix, and the headroom for downstream value-added growth. On raw-material dependency, the bauxite import-to-domestic ratio reached 2.6:1 in 2025, with an import dependency of 72 percent versus 42 percent in 2015. On power-mix carbon intensity, average tCO2 per tonne aluminum was 11.2 tonnes nationally — substantially higher than peer countries (Norway's Hydro at 2.2 tonnes, Alcoa at 5.8 tonnes). On value-added headroom, China's aluminum processing portfolio remains skewed toward mid-low-end construction materials, with imports still significant in high-end aerospace plate, precision automotive sheet, and premium battery foil.

Heading into 2026, several structural shifts deserve heightened attention. First, the EU Carbon Border Adjustment Mechanism (CBAM) entered the formal collection phase for aluminum from January 2026, charging based on embedded carbon at rates equivalent to EUR 50-90 per tonne initially, with annual upward revaluation tracking EU ETS prices. Second, Guinea raised its bauxite export tariff again in mid-2025, from 15 to 17 percent, and 2026 signals point to potential quota management — disrupting upstream alumina supply globally. Third, battery-foil demand entered super-linear growth as China's NEV sales broke 17 million units — Chinese battery-foil output rose from only 320,000 tonnes in 2022 to 720,000 tonnes in 2025, with continued acceleration in H1 2026. Fourth, the national carbon market (CEA) formally extended to the electrolytic-aluminum sector — a milestone in China's carbon-market expansion, imposing direct additional compliance costs on coal-power aluminum producers.

From an industrial-organization perspective, the Chinese aluminum sector further consolidated in 2025. The top eight players — Chalco, China Hongqiao, Yunnan Aluminium, Shenhuo, Tianshan Aluminium, Nanshan Aluminum, East Hope Group, and Xinfa Group — together held 81.4 percent of electrolytic-aluminum capacity, up 10 points from 71 percent in 2020. Green aluminum, clean aluminum, and recycled aluminum became the three most-watched capital-market themes. Yunnan, Inner Mongolia, and Xinjiang together carried 46 percent of electrolytic-aluminum capacity, with Yunnan hydropower-aluminum exceeding 15 percent — a configuration unimaginable a decade ago. By capacity scale, eight enterprises produce over 1 million tonnes annually, six 0.5-1 million tonnes, eleven 0.2-0.5 million tonnes, and fifteen below 0.2 million tonnes — head-end concentration still has room to climb.

In downstream processing, China produced 16 million tonnes of aluminum sheet & plate, 5.8 million tonnes of aluminum foil, and 24 million tonnes of aluminum extrusion in 2025 — 46 million tonnes combined, comprising 82 percent of electrolytic-aluminum consumption. Automotive sheet, battery foil, PV frames, rail-transit extrusions, can stock, aerospace alloy, consumer-electronics alloy, and home-appliance alloy form the "eight downstream pillars" of deep aluminum processing. From a demand-mix angle, traditional construction aluminum's share fell from 35 percent in 2015 to 26 percent in 2025, while the combined share of NEV aluminum, battery aluminum, PV aluminum, and rail-transit aluminum rose from 11 to 29 percent in the same period.

From the demand side, NEVs were the strongest growth engine for Chinese aluminum consumption in 2025. China sold 17.20 million NEVs in 2025, with average aluminum consumption per vehicle of 195 kg, equating to total NEV-aluminum demand of around 3.35 million tonnes; by 2030, NEV sales are expected to reach 30 million units with average aluminum per vehicle of 250 kg, lifting NEV-aluminum demand to 7.50 million tonnes — more than doubling in five years. PV frames are the other powerful engine: 2025 global PV new-build capacity reached 650 GW, requiring 3.20 million tonnes of aluminum frames; by 2030, global PV new-build is expected to exceed 1,000 GW, lifting frame demand to 5 million tonnes.

Global supply-demand fundamentals show "modest deficit" in 2025 — global primary supply 72.50 million tonnes, global primary consumption 72.80 million tonnes, gap of about 300,000 tonnes. The gap was filled by global stocks (LME warrants, SHFE warrants, and bonded inventories), with LME aluminum stocks falling 130,000 tonnes year-on-year to 560,000 tonnes by year-end 2025, SHFE stocks down 90,000 tonnes to 180,000 tonnes, and Chinese bonded stocks down 110,000 tonnes to 220,000 tonnes. Globally visible aluminum stocks total approximately 960,000 tonnes — equivalent to five days of global consumption, approaching the operational floor.

China's global position can also be characterized through trade flows. In 2025, China was a net exporter of approximately 1.12 million tonnes of primary aluminum, with breakdown across aluminum alloy ingots 580,000 tonnes, wire & rod 210,000 tonnes, sheet & plate 240,000 tonnes, foil 610,000 tonnes, and extrusions 260,000 tonnes — combined Chinese aluminum-product trade flowing outward at approximately 1.90 million tonnes (net of the alloy-ingot import and re-export double-counting). China exported 610,000 tonnes of aluminum products to the EU, 320,000 tonnes to North America, 480,000 tonnes to Southeast Asia, 210,000 tonnes to the Middle East and Africa, 180,000 tonnes to Russia/CIS. China's aluminum export structure faces multiple external frictions in 2026 — CBAM, Indian anti-dumping, Turkish anti-dumping, Mexican anti-dumping — among others.

Looking ahead, the remainder of this report analyzes the 2025-2026 Chinese aluminum landscape across eight dimensions — mining sources, refining, electrolysis, processing, recycling, pricing, policy, and risk — and closes with our research institute's three-to-five-year directional judgment. We deliberately maintain the research-institute register — refraining from preset conclusions and single-metric emphasis, instead building a complete, multi-dimensional, verifiable picture of Chinese aluminum from the actual operating reality at every node of the value chain.

2. Upstream — Guinea Boké port, Australian Weipa, Indonesian Kalimantan

The starting point of the aluminum value chain is bauxite. China is the world's largest bauxite consumer but its domestic resource endowment is not abundant — proven reserves are around 1 billion tonnes, dominated by high-silicon diaspore with an average Al-Si ratio of only 5-7, well below the global average of 12-15. Mines are scattered across Xiaoyi (Shanxi), Sanmenxia (Henan), northern Guizhou, Pingguo (Guangxi), and Wenshan (Yunnan); single-mine scale is limited and mining costs continue to rise. Domestic raw bauxite output in 2025 fell to 60.50 million tonnes, down nearly 40 percent from the 2018 peak — some mines forced to halt operation due to ecological restoration and tighter county-level safety oversight.

What fills the gap is highly concentrated imported bauxite. China imported 158 million tonnes of bauxite in 2025: Guinea alone supplied 110.50 million tonnes (70 percent), Australia 35 million tonnes (22 percent), and Indonesia 9.40 million tonnes (6 percent) — together 98 percent. This extreme concentration is the single greatest geopolitical fragility in the 2025 aluminum value chain.

Guinea's bauxite comes mainly from the western Boké region, dominated by gibbsite — 45-50 percent alumina content, Al-Si ratio above 30 — the world's premier feedstock for the Bayer process. Chalco, China Hongqiao, East Hope, Yantai Zhongtai, TBEA, and SDIC Power all hold equity or joint ventures around Boké. The largest is SMB (Société Minière de Boké), the joint venture between China Hongqiao, Singapore-based Winning International, Guinea's UMS, and the Guinean state mining asset-management corporation; SMB mined 50 million tonnes in 2025, making it West Africa's largest open-pit bauxite operation. Chalco's wholly owned Boffa mine produced 22 million tonnes in 2025, with cargo shipped via the dedicated fleet from Kamsar deep-water port southeast of Conakry, taking 42 days to reach Yantai, Weifang, Dongying, and other northern Chinese ports.

Guinea has repeatedly tightened its bauxite export regime in 2024-2025. The export tariff was raised from 11 to 15 percent in September 2024; from 15 to 17 percent in August 2025; and in December 2025 Guinea's National Assembly passed the Mining Code Amendment, requiring matching alumina-refining capacity within Guinea before raw bauxite exports can continue from 2027. This series of policy actions directly pushed Q4 2025 international alumina spot prices higher — CIF China alumina briefly approached RMB 5,650 per tonne in December 2025, up 52 percent from the March 2025 trough. By Q1 2026, with alumina supply chains recovering and new Chinese alumina capacity coming online, prices retreated to around RMB 3,950.

International institutions rate Guinea's political risk at elevated levels. BloombergNEF's 2026 Global Bauxite Supply Risk Map flags Guinea as "high-sensitivity," ranking just behind Russia's Far East and the Democratic Republic of the Congo. Reuters reported in March 2026 that Guinea's ministry of mines is evaluating "million-tonne-level export quotas," aiming to compel foreign miners to accelerate domestic alumina-refinery construction. The Nikkei focused on Kamsar deep-water port capacity bottlenecks — actual 2025 throughput at 110 million tonnes is closing in on the designed cap of 130 million tonnes. Even absent further Guinean policy tightening, physical constraints may bind bauxite exports within two to three years.

Australia's bauxite comes from three principal regions: Rio Tinto's Weipa open-pit in Queensland, producing 34 million tonnes in 2025; Rio Tinto's Gove peninsula mine in the Northern Territory, producing 12 million tonnes; and Worsley in Western Australia (Southern Copper's holding), producing 15 million tonnes. Weipa ore has high gibbsite content and short shipping distance to Chinese ports (~21 days), but the unit cost is roughly 12 percent above Guinean ore. Australia's supply path is relatively stable — the Australian government has not imposed special restrictions on bauxite exports, although the 2023 trade friction with China saw sporadic export-approval delays, positioning Australia strategically as a "supplementary supplier" rather than the primary source.

Indonesia's bauxite comes from West Kalimantan, jointly developed by Indonesian state miner Antam together with Tianshan Aluminium, Qingdao Chalco, and East Hope. Notably, since President Jokowi's January 2023 decision to ban raw bauxite exports outright to incentivize domestic alumina refining, Indonesia's bauxite exports fell from a 2018 peak of 65 million tonnes to under 10 million tonnes by 2025; domestic alumina capacity meanwhile expanded from 2 to 7.80 million tonnes. Chinese-led Indonesian alumina projects — Tianshan Aluminium's Bintan 2 million tonnes, India's NALCO's 1 million tonnes at Mempawah, China Hongqiao with Winning International's 3 million tonnes at Galang Batang — are pillars of Indonesia's industrial policy. Indonesia's domestic alumina capacity is expected to break 15 million tonnes by 2028, with some flow back to Chinese electrolytic-aluminum smelters as feedstock.

To map the real upstream-source landscape of Chinese aluminum, one can search for alumina refineries, bauxite processors, and aluminum casting plants to identify the actual operating plants.

Among remaining domestic bauxite resources, the Xiaoyi area in Shanxi has built a relatively complete "mine-aluminum-power" integrated layout, but ore-grade decline is acute — the Al-Si ratio has fallen from 8.5 two decades ago to 5.2 in 2025, with alumina unit power and alkali consumption rising sharply. Sanmenxia in Henan is gradually depleting; some firms shifted to overseas mine operations. Guizhou Zunyi and Guangxi Pingguo face ecological red-line constraints on expansion. Wenshan in Yunnan serves as Yunnan Aluminium's domestic raw-material reserve, with the adjacent northern Vietnam bauxite zone also under long-term offtake. Overall, China's bauxite import dependency is expected to remain in the 68-72 percent range in 2026; import concentration shows no signs of easing, and Guinea will remain the most sensitive "chokepoint" in the global aluminum value chain in the near term.

The Al-Si ratio is the most direct indicator of bauxite grade. Guinean gibbsite ore typically scores 30-40, occasionally 50; Australian Weipa scores 22-28; Indonesian West Kalimantan scores 8-12 (low-grade gibbsite); Chinese mines at Xiaoyi, Sanmenxia, Zunyi, and Pingguo have fallen to 5-7, with some old Xiaoyi mines below 4. Each one-point drop in Al-Si ratio raises alumina alkali consumption, power consumption, red-mud production, and combined costs in tandem. This physical constraint means that even modest new domestic bauxite discoveries cannot reverse the structural reality that "premium resources are overseas, low-grade resources are domestic."

Logistics is an underestimated layer. From Kamsar in Guinea to northern Chinese ports, single-vessel transit is 40-45 days; Capesize freight averaged USD 18.5 per tonne in 2025 and Panamax USD 22.8 per tonne. In Q1 2026, Red Sea geopolitical risk and Suez Canal diversions via Singapore briefly pushed rates to USD 22-26 per tonne. Freight upticks fed directly through to CIF prices, raising the comprehensive raw-material cost at the smelter gate by approximately 4.8 percent in Q1 2026. Chalco, China Hongqiao, East Hope, and Tianshan Aluminium have signed three-to-five-year long-term contracts with COSCO Shipping, China Merchants Group, and Pacific Basin Shipping to lock in capacity against short-term volatility.

Bauxite grade and recovery rate determine the alumina refinery's process route and integrated cost. Guinean gibbsite is processed by low-temperature Bayer (140°C, 0.4 MPa) — unit alkali ~90 kg NaOH per tonne alumina, unit power ~280 kWh; Chinese diaspore demands high-temperature Bayer (240-260°C, 1.8-2.5 MPa) — unit alkali ~150 kg, unit power ~420 kWh. The gap means refineries using Guinean ore enjoy RMB 200-300 per tonne lower combined cost. New Chinese alumina projects almost universally plan around imported Guinean ore — Chalco Zhongzhou's 3 million tonnes Henan project, Huasheng Materials' 2 million tonnes Guangxi project, and Tianshan Aluminium's Bintan 2 million tonnes Indonesia project all exemplify this trend.

Bauxite price differences by country also bear noting. In 2025, Guinean Boké FOB was USD 42 per tonne and CIF China USD 76 — freight was 44.7 percent of CIF, among the highest freight-to-price ratios in major bulk minerals. Australian Weipa FOB was USD 48 and CIF China USD 82, with freight at 41.5 percent. Indonesian West Kalimantan FOB was USD 32 and CIF China USD 64, with freight at 50 percent — short shipping distance but lower unit price. Domestic Chinese bauxite at the mine gate sells for RMB 120-160 per tonne (USD 17-23) — well below imported, but requires additional land transport to the refinery.

Domestic exploration and development face severe practical constraints. Shanxi, Henan, Guizhou, and Guangxi — the four provinces with concentrated bauxite reserves — almost universally operate dual-license (exploration + mining) systems, with new-mine approval cycles of 3-5 years and ecological-restoration costs rising from 12 percent of mining costs two decades ago to 38 percent today. Yunnan Wenshan, given its location upstream of the Pearl River within an ecological-protection zone, faces dual constraints from the Ministry of Ecology and Environment and the Yunnan government on single-mine scale and annual mining quotas. Overall, Chinese domestic bauxite output is likely to stabilize at 55-60 million tonnes annually over the next five to seven years, with no prospect of returning to the 2018 peak.

For exploration and overseas reserve building, Chalco, China Hongqiao, East Hope, Shenhuo, and other major players have rolled out bauxite exploration or concessions in Sierra Leone, Liberia, Ghana, Republic of Congo, Angola, Turkey, Iran, Kazakhstan, Brazil, and Guyana. Turkey's Etibank bauxite, Brazil's Pará gibbsite, and Guyana's Linden gibbsite are next-tier potential primary supply sources. But overseas bauxite development is constrained by local infrastructure (ports, railways, power) bottlenecks and geopolitical risk — from discovery to stable supply typically requires 7-10 years.

International bauxite reserves data deserve refresh. The USGS 2026 Mineral Commodity Summaries report puts proven global bauxite reserves at 32 billion tonnes. Guinea 7.4 billion (23 percent), Australia 5.8 billion (18 percent), Vietnam 5.2 billion (16 percent), Brazil 2.6 billion (8 percent), India 3 billion (9 percent), Indonesia 1 billion (3 percent), Jamaica 0.7 billion (2 percent), Ghana 0.7 billion (2 percent), Russia 0.5 billion (2 percent), other 4 billion (12 percent), China domestic remaining ~1 billion (3 percent). Vietnam's bauxite reserves are large but the government strictly controls exports, requiring domestic alumina refining. The medium-to-long-term release pathway of Vietnam's resources is a future variable to watch.

3. Process barriers — Bayer, Hall-Héroult, anode carbon, and cell upscaling

Aluminum smelting rests on two cornerstone processes: Bayer (alumina from bauxite) and Hall-Héroult (aluminum from alumina via electrolysis). Both were developed in the 1880s and remain the cornerstones of global aluminum, with substantial process refinement across 140 years.

The Bayer process digests bauxite in concentrated NaOH solution under pressure — alumina dissolves as sodium aluminate while silicon, iron, and titanium impurities remain in red mud; the solution is diluted, seed-crystallized to precipitate gibbsite, then calcined to dehydrate to alumina powder. Key control parameters are temperature, pressure, alkali concentration, and Al-Si ratio. For high-grade Guinean gibbsite, digestion runs at 140°C and 0.4 MPa with alumina recovery above 92 percent. For Chinese diaspore, digestion requires 240-260°C and 1.8-2.5 MPa, with recovery only 85 percent and higher red-mud alkali content and worse consolidation properties.

Red-mud handling is China's century-long unresolved environmental issue. Every tonne of alumina produces 1.5-2 tonnes of red mud; nationally, cumulative red-mud inventory reached approximately 1.2 billion tonnes. Red mud is strongly alkaline (pH 12-13) with trace radioactive elements, iron, sodium, vanadium, and cobalt. The traditional disposal method is dam impoundment; following the 2008 Hungarian Ajka red-mud dam failure, China tightened design standards and monitoring requirements year after year. In 2025, Chalco, China Hongqiao, and East Hope formed a red-mud comprehensive utilization industry consortium — engineered pathways for cement blends, road bed fill, iron-ore recovery, and titanium-dioxide byproduct aluminum salt recovery — but truly consumable proportions are only 11 percent of annual output. Rising red-mud disposal costs are reverse-driving refiners toward higher-grade imported gibbsite.

Internationally, the most advanced red-mud comprehensive utilization research is conducted by Australia's CSIRO, Canada's CRIC, and Norway's SINTEF (within Norsk Hydro). SINTEF has developed a "red-mud iron, aluminum, sodium staged-recovery" process, capable of raising metal recovery to 62 percent. But the combined cost is around EUR 300 per tonne of red mud — still away from industrial economics. Chalco and China Hongqiao maintain technical exchanges with SINTEF and CSIRO but have not yet introduced industrial production lines.

Hall-Héroult electrolysis reduces alumina dissolved in molten cryolite (Na3AlF6) at 950°C; cell voltage 4.2-4.5 V; current injected via top carbon anodes and drained from the bottom cathode; dense aluminum melt collects at the cell bottom. Cell design current strength is the headline process-sophistication indicator. Cells of 160 kA in the 1980s produced ~1,000 tonnes per cell-year; 350-500 kA cells dominated the 2010s; 600 and 660 kA cells came online in the 2020s, with single-cell output of 3,800 tonnes per year.

Chalco's Lanzhou subsidiary commissioned a 660-kA cell in 2025 — China's largest, with the lowest specific energy: 12,500 kWh DC per tonne of aluminum, 7.4 percent below the industry average of 13,500. China Hongqiao's 600-kA cell upgrade jointly with Norsk Hydro saw specific energy fall to 12,800 kWh in the same year. Globally, Rusal's RA-550 cell (550 kA) commissioned in 2024 reaches 12,200 kWh — currently the lowest-energy carbon-anode cell. Rio Tinto and Alcoa's jointly developed AP60 smart cell (600 kA) reaches 12,700 kWh, deployed in Quebec. EGA's DX+ Ultra (620 kA) reaches 12,900 kWh — the mainstream choice for new electrolytic-aluminum plants in the Middle East and Africa.

Cell design also involves several critical detail factors: anode centering accuracy, magnetic-field compensation, cathode conductor materials, in-cell ledge management, and electrolyte chemistry (calcium fluoride, lithium fluoride, magnesium fluoride additives). The cathode conductor rod is a copper-steel composite — a critical component for reducing bottom-voltage drop; China Hongqiao and Baowu Steel jointly developed high-conductivity copper-steel composites now deployed at Shandong Qiaoji and Yunnan Wenshan Funing, saving ~0.8 percent per cell. Lithium fluoride additives can reduce electrolyte temperature by 20-30°C and unit power by 3.2 percent, but global supply is tight — ton-prices rose from RMB 100,000 in 2021 to RMB 380,000 in 2025, constraining large-scale adoption.

Anode carbon is another hidden cost core of electrolytic aluminum. Each tonne of aluminum consumes ~410 kg of prebaked anode; 2025 national prebaked anode output was ~18 million tonnes, with revenue close to RMB 100 billion. Prebaked anodes use petroleum coke and coal-tar pitch as raw materials, kneaded, formed, and baked in special ring furnaces at 1,200°C for 20-26 days. China Hongqiao's subsidiary Sunstone Development, Shandong Qili New Materials, and Guangxi Qiangqiang Carbon are the three largest domestic anode-carbon producers. Rio Tinto's AP60 smart cell at Quebec uses a proprietary anode formula that compresses consumption to 380 kg per tonne aluminum. The primary feedstock for prebaked anodes — petroleum coke — is highly concentrated among Saudi Aramco, ADNOC, Reliance Industries, Sinopec, and CNPC's specific refineries — global output around 80 million tonnes annually, supporting roughly 2 million tonnes of incremental global aluminum production — a hidden ceiling on anode capacity.

The most-anticipated process revolution is inert anode technology. The ELYSIS joint venture (Rio Tinto and Alcoa) commissioned its first 450-kA inert-anode pilot at Quebec in 2024 — emitting oxygen rather than CO2 per tonne aluminum, theoretically saving 8 tonnes of CO2 per tonne aluminum. ELYSIS targets 600-kA inert-anode commercial operation by 2027. Rusal also runs an inert-anode pilot cell (140 kA) at Sayanogorsk and in 2024 produced the world's first commercial "zero-carbon aluminum" samples, ordered in small lots by BMW, Audi, and Daimler for premium-model bodies. Chalco, China Hongqiao, and East Hope are all tracking inert-anode pathways, but commercial windows are judged to be post-2030. The core technical challenge is the long-life and corrosion-resistance of ceramic-metal composite anodes — the 950°C strongly oxidative cell environment is harsh. ELYSIS's current composite anode has a theoretical life of 1,500 hours, still short of the 8,000-hour commercial requirement.

"Green-power replacement" is the third axis of Chinese electrolytic-aluminum process upgrade. Five green-power aluminum hubs have taken shape: Yunnan hydropower aluminum, Inner Mongolia wind-PV aluminum, Xinjiang PV aluminum, Sichuan curtailment-water aluminum, and Qinghai PV aluminum. Yunnan Aluminium's Wenshan, Zhaotong, Honghe, and Wenshan Funing smelters operate with 88 percent hydropower; tCO2 per tonne aluminum was 3.5 tonnes — about 80 percent below the national average of 16. Green-power aluminum has thus entered the low-carbon supplier whitelists of Western automakers (BMW iVision, Mercedes-Benz EQ, Volkswagen ID), as well as Apple, Google, and Amazon's procurement frameworks. The premium has stabilized at RMB 300-500 per tonne.

Engineering details of green-power replacement also deserve unpacking. Yunnan hydropower aluminum is supplied mainly from Jinsha River cascade (Baihetan, Wudongde, Xiangjiaba, Xiluodu) and Lancang River cascade (Xiaowan, Nuozhadu, Huangdeng, Miaowei). Yunnan's 2025 dispatchable hydropower output was ~280 billion kWh; after residential, other industrial, and West-to-East transmission to Guangdong, the supply available for electrolytic aluminum is ~72 billion kWh — supporting capacity around 5.30 million tonnes at 13,500 kWh DC per tonne. With Yunnan's actual aluminum capacity at 6.20 million tonnes, around 1 million tonnes still rely on supplementary thermal — Yunnan's green-power aluminum is not strictly "zero-carbon." For the Inner Mongolia wind-PV pathway, Tongliao and Ordos hubs have brought wind-PV-storage projects online, but intermittency necessitates baseload thermal — wind-PV self-share is typically 30-40 percent.

Cell life management is the operational-economics heart of an electrolytic-aluminum plant. A 600-kA cell has design life ~2,400-2,800 days (6-7.5 years); shell deformation, lining erosion, conductor-rod aging, and magnetic-compensation device failure determine cell life. Life extension hinges on stable operating parameters (current, temperature, electrolyte composition all confined to narrow bands). Chalco, China Hongqiao, and East Hope's cell average life is 2,800-3,200 days — above the international average. A 500,000-tonne plant typically operates 280-320 cells in seven-year rotation cycles. Each cell relining costs RMB 3-4 million; annual relining bill across a plant runs RMB 130-170 million.

Safety management is another notable process detail. Major safety risks include cell aluminum spill (electrolyte or aluminum melt breakthrough), pneumatic explosion (aluminum melt contacting cooling water), cryolite-dust explosion, casting-shop mechanical injury, and high-voltage electrical hazards. Aluminum-spill frequency averages ~0.2 incidents per 10,000-tonne-year capacity; per-incident loss (equipment, melt, downtime) runs RMB 3-5 million. The Chinese electrolytic-aluminum sector has cut spill frequency by ~60 percent versus 2015, by rolling out intelligent monitoring of cell current, temperature, electrolyte composition, shell stress, and melt height. Core hardware is distributed fiber-optic sensors and infrared thermal cameras, supplied by ABB, Siemens, Honeywell, Schneider Electric, plus domestic players such as China Aluminum Smart and Xinjiang Joinworld subsidiaries.

Beyond cells, casting and aluminum-melt transport entail their own process barriers. The casting shop typically includes continuous (DC) casting, vertical semi-continuous casting, and horizontal semi-continuous casting. DC casting yields slab and billet; semi-continuous yields initial automotive-sheet slab. On melt transport, "short-distance integrated" smelter-processor co-location matters — per additional 50 km of melt transport, ~2.8 kWh per tonne is lost in heat; beyond 200 km, melt must be re-cast as ingot and shipped, costing another ~200 kWh per tonne. Yunnan, Xinjiang, and Inner Mongolia electrolytic-aluminum clusters have all planned "smelter-plus-deep-processing" integrated parks.

Waste-heat recovery and water cycling are quieter process details. Cell operating temperature is 950°C, per-cell heat dissipation ~400 kW, cumulative national waste heat ~40 billion kWh per year. Economically recoverable is ~12-18 percent, used for heating, casting-shop preheat, and melt insulation. Chalco's Qinghai Xihai Aluminium, Inner Mongolia Ulanqab Dalat Aluminium, and Xinjiang Changji Aluminium all run waste-heat recovery systems, saving ~120 million kWh per plant per year. On water, cell fluoride scrubbing, cathode cooling, and melt cooling consume substantial circulating water — ~30 tonnes per tonne aluminum, with reuse rate above 95 percent — a "water-sensitive" feature.

Fluoride emissions are also worth highlighting. Cryolite inevitably volatilizes at 950°C, releasing gaseous HF and particulate fluorides (NaF, AlF3). Chinese standard GB 25465-2010 caps fluoride emission concentration at 3 mg/m3. Modern cells run wet or dry fluoride scrubbers — dry absorption efficiency reaches 99.5 percent, with recovered fluorides reused as electrolyte additives. Each 1-percent rise in fluoride recovery saves about RMB 1.5 per tonne in electrolyte costs.

Workforce structure also exhibits distinctive features. A modern 500,000-tonne plant employs 600-800 people: 30 percent cell operators, 20 percent casting-shop operators, 20 percent maintenance and safety, 15 percent technical and management, 15 percent auxiliary. The core cell-operator skill is "cell-watching" — observing current, voltage, temperature, electrolyte composition, and melt height to judge cell condition and adjust parameters. A qualified cell-watcher typically needs 2-3 years of on-site training; the high-temperature environment is a long-term health challenge. The industry is transitioning from labor-intensive to intelligence-assisted operation — through cell-monitoring AI, auto-feeding, and auto-tapping systems, per-capita output rose from 600 tonnes in 2015 to 850 tonnes in 2025.

To map the actual operating state of every electrolytic-aluminum plant, alumina refinery, and anode-carbon plant, refer to electrolytic aluminum plants, prebaked anode plants, anode carbon plants, and aluminum-ingot smelters for the factory map.

4. Major players — Chalco, Yunnan Aluminium, Tianshan, Shenhuo, Nanshan, Mingtai, Hongqiao

Chalco (601600.SH) posted 2025 revenue of RMB 268.5 billion and net profit of RMB 14.3 billion, with alumina output 17.40 million tonnes and primary aluminum 7.60 million tonnes. Chalco is the most vertically integrated national-team player — from Guinea Boffa bauxite to alumina refineries in Shanxi, Henan, Guangxi, and Guizhou, electrolytic-aluminum smelters in Qinghai, Guizhou, Inner Mongolia, Guangxi, Shanxi, and Lanzhou, and alloy-processing plants in Luoyang, Guiyang, and Xining; the chain extends down into rare-earth aluminum alloys, recycled aluminum, aluminum powder, and aluminum foil. Bauxite self-sufficiency reaches 93 percent and alumina self-sufficiency 100 percent — the only Chinese electrolytic-aluminum producer not dependent on external alumina purchase. Chalco's Guangxi subsidiary Pingguo Aluminium is China's largest diaspore-processing base, in operation for nearly forty years; the Shanxi subsidiary Xingxian alumina project (3 million tonnes/year) reached full ramp in 2025 — the largest single-line alumina refinery in China.

Chalco's full-chain strategy is also reflected in its subsidiaries: Guizhou Wuchuan Aluminium (bauxite), Guiyang Aluminium-Magnesium Design Institute (process design and equipment), Shenyang Aluminium-Magnesium Design Institute (cell and casting-shop design), Changsha Nonferrous Metals Research Institute (cell process research), and Qingdao Aluminium-Magnesium Technology (lightweight alloy R&D) — a closed loop of "mining-process-equipment-design-research." This loop makes Chalco not only a producer but also the core equipment provider for Chinese electrolytic-aluminum process design — domestic cell, casting-shop, and anode-baking-furnace designs are predominantly from Chalco's Guiyang and Shenyang institutes.

China Hongqiao (1378.HK) posted 2025 revenue of RMB 156.4 billion and net profit of RMB 23 billion, with primary aluminum 6.46 million tonnes and alumina 18.90 million tonnes — overtaking Chalco as the world's largest alumina producer. Hongqiao was historically based at Weiqiao Town in Binzhou, Shandong, hosting Asia's largest single-site aluminum cluster, but since 2020 has gradually relocated 4 million tonnes of Shandong capacity to Wenshan Funing in Yunnan; by 2025 Yunnan capacity reached 4.20 million tonnes and Shandong residual capacity 2.20 million tonnes. Hongqiao's subsidiary Sunstone Development is China's largest independent anode-carbon producer. Worth noting: Hongqiao's controlling shareholder Weiqiao Pioneering Group also pursues aggressive overseas positioning — beyond Guinea SMB, equity stakes in Indonesia Galang Batang alumina and Malaysia Sime Cement. In 2025, Hongqiao began clearly disclosing its "green-power aluminum transition pathway" to international investors; in June 2026, it secured ASI "performance standard" full certification — the second Chinese smelter to do so.

Hongqiao's operational efficiency and upstream integration are differentiators. The Shandong Qiaoji cluster was long the lowest-cost Chinese electrolytic-aluminum plant (combined cost ~RMB 14,800 per tonne); after relocation, Yunnan plants reach ~RMB 15,200 per tonne (including hydropower purchase and cross-province transport). Hongqiao's Sunstone Development (603612.SH) is the largest independent anode-carbon producer in China, with capacity of 2.80 million tonnes, supplying Hongqiao, Yunnan Aluminium, East Hope, Shenhuo, and others. Sunstone reported 2025 revenue of RMB 10.2 billion and net profit of RMB 1.1 billion — the absolute leader in anode-carbon.

Yunnan Aluminium (000807.SZ) reported 2025 revenue of RMB 56.1 billion and net profit of RMB 6.8 billion, with electrolytic-aluminum output 3.05 million tonnes. Its four smelters at Wenshan, Zhaotong, Honghe, and Wenshan Funing run on 88 percent hydropower — China's "green aluminum benchmark." Deep-processing capacity is 1.50 million tonnes — Kunming foil, Wenshan billet, and Honghe sheet & plate have all moved into NEV battery foil, automotive sheet, and can stock. Yunnan Aluminium's controlling shareholder, China Copper (under Chalco Group), forms a "Group-internal division of labor" with Chalco — Chalco focuses on full chain and high-end alloys, Yunnan Aluminium on green-power aluminum and mid-end deep processing. 2025 per-tonne net profit was RMB 2,200, third in the industry — only behind China Hongqiao (RMB 2,600) and Tianshan Aluminium (RMB 2,800).

Tianshan Aluminium (002532.SZ) reported 2025 revenue of RMB 32.8 billion and net profit of RMB 3.4 billion, with primary aluminum 1.20 million tonnes and alumina 2.50 million tonnes. Tianshan Aluminium is the only private firm running coal-power-aluminum integration in Xinjiang, with its Shihezi smelter fully on captive coal power and tCO2 per tonne aluminum power cost at RMB 2,700, well below the industry average of RMB 3,800. It also aggressively positions in Indonesia's Sulawesi alumina project (designed annual output 2 million tonnes), commissioned in H2 2025 — opening up Indonesian green-power aluminum export. Xinjiang Fukang Phase 1 (600,000 tonnes electrolytic aluminum) was commissioned in September 2025, lifting total Tianshan capacity to 1.80 million tonnes. Worth noting: Tianshan is the most profitable Chinese electrolytic-aluminum producer — 2025 combined cost was RMB 14,800 per tonne, per-tonne net profit was RMB 2,800.

Shenhuo (000933.SZ) reported 2025 revenue of RMB 39.5 billion and net profit of RMB 4.2 billion, with primary aluminum 1.70 million tonnes — 850,000 tonnes at Shangqiu (Henan) and 900,000 tonnes at Zhaotong (Yunnan). Shenhuo is the archetype of coal-power-aluminum integration, with subsidiary Yongcheng Coal Power providing in-house electricity to the Shangqiu smelter. Yunnan Shenhuo's smelter is the industry's "hydropower replacement" benchmark — since 2020's Henan-to-Yunnan relocation, per-tonne combined power cost has fallen 31 percent. The most recent move is the Phase-2 Henan Shangqiu disassembly-and-relocation to Yunnan Zhaotong, launched in Q4 2025; the remaining 400,000 tonnes will complete the migration by 2027. Post-relocation, Yunnan Shenhuo will reach 1.30 million tonnes and Shangqiu will go to zero — making Shenhuo Yunnan's third-largest smelter after China Hongqiao and Yunnan Aluminium.

Nanshan Aluminum (600219.SH) reported 2025 revenue of RMB 32 billion and net profit of RMB 2.1 billion, with primary aluminum 820,000 tonnes and aluminum sheet & plate 950,000 tonnes — including aerospace plate 65,000 tonnes, automotive sheet 250,000 tonnes, and battery foil 50,000 tonnes. Nanshan is the deepest domestic player in the "aerospace alloy" segment, having become an approved supplier to COMAC's C919, AVIC, Airbus China assembly, and Boeing China aftermarket. Nanshan's Indonesia Bintan alumina project (designed 2 million tonnes/year) has commissioned 1 million tonnes, supplying internal smelters. Aerospace alloy plate is the key product driving import-substitution — China's main aerospace structural alloy plate had been almost entirely supplied by Alcoa, Constellium, and Rio Tinto. Since breaking through the 7050-series large-thickness aerospace alloy plate in 2018, Nanshan has progressively achieved C919 and ARJ21 main-structure-alloy domestic substitution.

Mingtai Aluminium (601677.SH) reported 2025 revenue of RMB 30.5 billion. It is one of the largest independent sheet-and-plate processors in China, with 1.20 million tonnes annual capacity, battery-foil capacity of 80,000 tonnes, and can stock 250,000 tonnes. Mingtai does not run electrolytic aluminum — focused entirely on downstream processing, signing long-term supply contracts with China Hongqiao, Yunnan Aluminium, and Shenhuo. The recent priority is breakthrough in automotive sheet — its domestic 6111- and 6022-series alloy sheets are now placed with Geely, Great Wall, Changan, GAC, and Chery's premium models. By 2027, Mingtai automotive sheet output is expected to break 400,000 tonnes, becoming the second-largest domestic supplier behind Novelis China.

Dingsheng New Materials (603876.SH) is the global leader in battery foil, with 2025 revenue of RMB 18 billion, net profit RMB 1.3 billion, and battery-foil output of 620,000 tonnes — global market share around 43 percent. Dingsheng operates battery-foil bases in Zhenjiang (Jiangsu), Hangjin (Inner Mongolia), and Huabei (Anhui), with overseas processing plants under construction in Germany, Turkey, and India, targeting local procurement by European, Turkish, and Indian battery makers. Core customers include CATL, BYD, LG Energy Solution, SK On, Samsung SDI, Tesla in-house cells, EVE Energy, and Sunwoda.

Huafon Aluminium (601702.SH) reported 2025 revenue of RMB 9.2 billion — the hidden champion in automotive-sheet segment. Its mainstay 5xxx-series automotive sheet (for body inner panels and battery-pack housings) and 3xxx-series foil (battery and AC foil) are key drivers of 5xxx-series automotive-sheet substitution. Clients include BMW China, Mercedes-Benz China, Audi-FAW, Tesla China (partial import substitution), NIO, Li Auto, and AITO.

Asia-Pacific Group (002540.SZ) reported 2025 revenue of RMB 8.6 billion — one of China's largest extrusion-only firms, products spanning architectural profiles, industrial extrusions, and automotive-alloy structures. It runs three extrusion bases in Nantong, Foshan, and Ordos with combined annual capacity of ~520,000 tonnes. Wanshun New Materials (300057.SZ) reported 2025 revenue of RMB 7.8 billion with foil output of 460,000 tonnes — a comprehensive foil producer covering battery foil, household, pharma, and packaging. Yongzhen Technology (603381.SH) is the lead PV-frame supplier, with 2025 revenue of RMB 5.8 billion and PV-frame output of ~200,000 tonnes, supplying LONGi, JA Solar, JinkoSolar, Trina Solar, Canadian Solar, and others.

Internationally, Rio Tinto's 2025 aluminum-segment EBITDA was USD 4.2 billion, primary output 3.34 million tonnes, bauxite 60.50 million tonnes, alumina 7.60 million tonnes. Rio Tinto's Quebec AP60 smart cell represents global state-of-the-art carbon-anode electrolysis. Rio Tinto Aluminium's core strength is full-chain upstream integration — Weipa and Gove in Australia, CBG in Guinea (joint venture), plus 11 electrolytic-aluminum plants across five countries (Canada, Australia, Iceland, New Zealand, Oman). Rio Tinto's average per-tonne carbon emission is ~2.8 tonnes — among the lowest of major global aluminum producers.

Alcoa Corporation reported 2025 revenue of USD 12.1 billion, primary aluminum 2.30 million tonnes, and alumina 9.85 million tonnes — global #2 alumina. Alcoa's Australian AWAC joint-venture (with Southern Copper/Saudi Mining) is the world's largest single alumina-refinery consortium. Alcoa's partnership in ELYSIS is a core strategic chip for the "low-carbon aluminum" future. Alcoa's 2025 strategy prioritizes AWAC capacity optimization and red-mud comprehensive use. AWAC operates four refineries in Western Australia, Queensland, and the Northern Territory (Pinjarra, Wagerup, Kwinana, Worsley) with combined ~14.20 million tonnes annual capacity. Alcoa also holds key positions in Juruti (Brazil) and Sangaredi (Guinea, via CBG). Alcoa's global-chain layout — versus Chalco/Hongqiao's "China-base + overseas extension" — exemplifies the "global layout + global sales" paradigm.

Rio Tinto's international integration includes the not-to-be-overlooked ELYSIS joint venture, holding inert-anode core IP, the global aluminum industry's most-promising route to commercial "zero-carbon" production. ELYSIS — held by Rio Tinto (40 percent), Alcoa (40 percent), Canadian government (10 percent), and Quebec government (10 percent) — has invested approximately CAD 900 million in R&D from 2023-2025. Its first inert-anode 450-kA pilot line was commissioned at Saguenay (Quebec) in 2024, with 600-kA commercial operation targeted for 2027. If ELYSIS delivers commercial scale on schedule, the global aluminum industry enters the "zero-carbon" era, with carbon intensity falling below 0.5 tonnes CO2 per tonne aluminum — far below today's best of 2-3 tonnes.

Norsk Hydro reported 2025 revenue of NOK 218 billion, primary aluminum 2.20 million tonnes, recycled aluminum 650,000 tonnes; the HyForge low-carbon aluminum brand has become Europe's preferred automotive-sheet choice. Hydro's electrolytic aluminum is mostly powered by Norwegian hydropower, with tCO2 per tonne aluminum of ~2.2 — the lowest globally. Hydro Extruded Solutions is Europe's largest extrusion firm, supplying European NEV industry, rail, and construction.

Hindalco / Aditya Birla Group reported 2025 revenue of INR 2.31 trillion (~USD 27.7 billion), primary aluminum 1.36 million tonnes, alumina 3.95 million tonnes. Hindalco's subsidiary Novelis is the world's largest can-stock supplier (4.60 million tonnes annual capacity), supplying Coca-Cola, PepsiCo, Nestlé, AB InBev, Carlsberg, Heineken, and other global beverage giants.

EGA — UAE state-owned — produces 2.70 million tonnes primary aluminum annually on natural-gas-powered electricity, the Middle East's only integrated aluminum giant. EGA owns Guinea GAC bauxite (14 million tonnes/year output) and UAE Al Taweelah alumina refinery (2.60 million tonnes/year). Per-tonne CO2 is ~9.5 tonnes — lower than coal but higher than hydropower.

Rusal produced 3.96 million tonnes primary aluminum in 2025, almost entirely from Siberian hydropower (Sayano-Shushenskaya, Krasnoyarsk, Bratsk). Rusal has one of the lowest per-tonne CO2 (~3.2 tonnes) globally but, under international sanctions, has progressively shifted sales focus from Western markets to China, the Middle East, and India during 2022-2025.

Vedanta Aluminium produced 2.42 million tonnes primary aluminum in 2025, with main bases at Jharsuguda (Odisha) and BALCO (Chhattisgarh). Vedanta runs predominantly coal-power, with tCO2 of ~13.8 tonnes — among the highest globally. But Indian domestic aluminum consumption is expanding rapidly (2025 Indian aluminum consumption broke 4.50 million tonnes, expected to break 8 million by 2030), so Vedanta's pace of expansion outruns cost pressure.

Brazil, Canada, Iceland, Oman, Bahrain, Mozambique also host distinctive electrolytic-aluminum operations. Brazil's Albras produces 460,000 tonnes. Canada Rio Tinto BC Works (Kitimat) — hydropower — produces 450,000 tonnes with ~1.8 tonnes per-tonne CO2 (one of the lowest globally). Iceland's Alcoa Fjarðaál, Rio Tinto ISAL, and Century Aluminium Norðurál together produce 820,000 tonnes, all running on Icelandic geothermal and hydropower with ~2.1 tonnes CO2. Oman Sohar Aluminium produces 410,000 tonnes on Omani natural gas. Bahrain ALBA produces 1.60 million tonnes — the Middle East's second-largest. Mozambique Mozal (South African Gold Group subsidiary) produces 600,000 tonnes, drawing from Cahora Bassa hydropower.

Among Chinese non-listed but consequential "hidden majors," East Hope Group runs the largest scale — ~2.10 million tonnes total electrolytic-aluminum capacity across Xinjiang Changji, Ningxia Zhongwei, Inner Mongolia Tongliao, and Qinghai Xining. East Hope's "coal-power-aluminum integration" strategy from founder Liu Yongxing has made it one of China's lowest-cost private operators. Xinfa Group runs ~1.80 million tonnes across Shandong Liaocheng, Xinjiang Changji, and Ningxia Zhongwei — Shandong's second-largest electrolytic-aluminum producer after Weiqiao (Hongqiao). Shandong Nanshan Group's parent retains some non-listed electrolytic-aluminum capacity beyond Nanshan Aluminum (listed).

Beyond these, several "small but specialized" focused players are noteworthy: Xinjiang Joinworld (600888.SH) leads on high-purity aluminum, electronic capacitor foil, and capacitor foil — 2025 revenue RMB 26 billion, top market share globally in electronic capacitor foil. Dongyangguang Aluminium (600673.SH) focuses on capacitor and formed foil — 2025 revenue RMB 9.8 billion, top-three globally. Dingsheng New Materials, Huafon Aluminium, Asia-Pacific Group, Wanshun New Materials, Yongzhen Technology — these specialized players collectively form the "long-tail force" in China's aluminum value chain.

To survey the players' map, see aluminum die-casting plants, aluminum alloy profile plants, aluminum sheet & plate plants, and aluminum billet plants — including both listed-company plants and many "hidden majors" not yet listed.

5. The 45-million-tonne ceiling — hydropower, wind-PV, and coal-power aluminum

The 45-million-tonne electrolytic-aluminum ceiling enacted by NDRC in 2017 is uniquely the "hard constraint" among Chinese metal-smelting sectors — no new approvals, no equivalent swaps to expand totals, only inter-firm transfers or inter-regional relocations of existing capacity. After eight years, this ceiling has become the strongest gravitational line shaping China's aluminum geography.

At year-end 2025, national operating capacity stood at 44.20 million tonnes; total built capacity at 44.56 million tonnes — leaving just 440,000 tonnes of headroom. Any firm wishing to expand electrolytic-aluminum operations must purchase "capacity indicators" from another firm — these indicators have become a quasi-option asset. 2025's per-10,000-tonne indicator trading price was approximately RMB 200 million, four times the 2018 trough of RMB 50 million.

Under ceiling pressure, electrolytic-aluminum capacity has been physically reshuffled. Between 2017 and 2025, capacity in Shandong, Henan, Xinjiang, Inner Mongolia, Qinghai, Guizhou, Guangxi, Gansu, and Ningxia contracted by ~6 million tonnes; "green-rich" regions like Yunnan, Sichuan, Inner Mongolia Tongliao, and Guangxi Chongzuo absorbed an equivalent of ~9 million tonnes of relocated capacity. Yunnan took the largest share — 2025 operating capacity reached 6.20 million tonnes versus 280,000 in 2017 (a 21-fold increase). Inner Mongolia Tongliao's wind-PV aluminum demonstration base reaches 1.20 million tonnes; Sichuan Ganzi's PV-aluminum demonstration base reaches 600,000 tonnes.

In the hydropower-aluminum lane, Yunnan is the leading actor. Yunnan Aluminium, China Hongqiao, Shenhuo, and East Hope have all built large smelter clusters there. Yunnan electrolytic aluminum draws primarily on the Lancang River, Jinsha River, and Nu River cascades, with annual generation of ~320 billion kWh, accounting for 76 percent of provincial generation. But hydropower aluminum is not the whole answer to "low-carbon" — Yunnan suffered severe summer drought in 2023, with Jinsha River cascade output falling 31 percent. China Hongqiao's Yunnan Wenshan smelter had to cut 200,000 tonnes of production; Yunnan Aluminium triggered peak-shaving plans. By 2025, Yunnan hydropower volatility had fallen 12 percent from 2021, mainly due to Wudongde and Baihetan operating at full capacity. But single-source hydropower dependency remains an unresolved "congenital risk."

In the wind-PV-aluminum lane, Inner Mongolia Tongliao and Ordos lead nationally. The Tongliao Horqin Sand Land PV base plus Zhalu Banner wind-power base has spawned the wind-PV-storage + electrolytic-aluminum integrated park at Holingol. Chalco's Baotou Aluminium, East Hope's Tongliao Huomei Aluminium, and Huomei Hongjun Aluminium together carry 1.20 million tonnes of wind-PV aluminum, with wind-PV-storage self-share at 32 percent — balance still purchased from Mengdong Grid. Ordos's Junggar Banner PV-aluminum project (China Hongqiao with Shaanxi Energy joint venture) reached trial operation in Q3 2025, with per-tonne CO2 of ~6.2 tonnes — between hydropower aluminum and coal-power aluminum.

In the coal-power-aluminum lane, Xinjiang and Shandong stand together. Xinjiang electrolytic aluminum — represented by Tianshan, East Hope, Xinfa Group, and Chalco Changji — relies on coal-rich Turpan Basin and Tianshan Northern Slope captive coal-power, with per-tonne power cost of RMB 2,500-3,100, but per-tonne carbon at ~16 tonnes. The biggest challenge for Xinjiang plants is the dual squeeze from EU CBAM and Chinese national carbon market. Shandong's veterans Weiqiao, Xinfa, and Yantai Nanshan have long been the coal-power-aluminum heavyweights, but capacity has steadily relocated since 2020. By 2025, Shandong's operating electrolytic-aluminum capacity dropped from the 2017 peak of 8.90 million tonnes to ~4.40 million tonnes.

Green-power aluminum, clean aluminum, and coal-power aluminum have diverged in price. The 2025 green (hydropower) premium over coal-power aluminum stabilized at RMB 300-500 per tonne. Entering 2026, with CBAM landing and the carbon market expanding, green premium widened to RMB 600-800 per tonne; some EU-bound green-aluminum orders fetch premiums of RMB 1,100. For the same electrolytic-aluminum tonnage, the Yunnan-versus-Xinjiang per-tonne annual net profit gap can reach RMB 400-700 — the strongest economic driver for relocation.

Three smaller but symbolically important green-power hubs deserve mention: Sichuan Ganzi, Qinghai Haixi, and Ningxia Zhongwei. Sichuan Ganzi's project, run by Chalco subsidiary Zhongfu Aluminium, has 600,000 tonnes scale, drawing on lower Jinsha and Yalong River cascades, hydropower share 94 percent and per-tonne CO2 of 2.8 tonnes — the lowest-carbon plant nationally. Qinghai Haixi's project, run by Chalco Qinghai with China Electric Power Investment, has 400,000 tonnes capacity, drawing on Qaidam Basin PV plus Yellow River cascades, per-tonne CO2 ~6.5 tonnes. Ningxia Zhongwei's project, run by Xinfa Group, was originally on Ningxia coal; in 2025 it launched a "green-power purchase contract" framework, targeting 50 percent green-power share by 2028.

Capacity relocation also involves substantial "indicator" market trades. By year-end 2025, cumulative national electrolytic-aluminum capacity-indicator transfers reached ~6.80 million tonnes, with cumulative deal value ~RMB 120 billion. Major sellers fall into six clusters: Qinghai, Gansu, Ningxia high-energy coal-power exits; depleted Henan Sanmenxia mines; Shandong "dual-control" forced exits; Guizhou ecological-constrained exits; Xinfa Group; and other phased-out smaller players. Buyers are concentrated in Yunnan Aluminium, Shenhuo, China Hongqiao, and Sichuan Ganzi Zhongfu Aluminium.

Green-aluminum certification systems are the "soft infrastructure" for relocation. China's "China Green Aluminium" voluntary certification by CNIA, the ASI "Performance Standard" plus "Chain of Custody" dual certification, EU Green Aluminium Origin Label, Apple's Low-Carbon Aluminum Supplier whitelist, and Tesla's Zero-Carbon Supply Chain framework — five layers of certification together form the "passport" for green-aluminum export. By year-end 2025, only China Hongqiao and Nanshan Aluminum have secured full ASI certification; Yunnan Aluminium and Tianshan Aluminium are in the certification pipeline.

Yunnan, Sichuan, Qinghai, Ningxia and other green-resource provinces' "carrying capacity" merits dedicated discussion. Yunnan installed hydropower is 110 GW, annual generation ~320 billion kWh, of which dispatchable hydropower (net of runoff-bound "must-generate" power) ~240 billion kWh; net of residential, commercial, other industrial, and West-to-East transmission to Guangdong (120 billion kWh promised), the available pool for electrolytic aluminum is ~82 billion kWh. At 13,500 kWh per tonne DC, Yunnan's "theoretical carrying ceiling" is ~6.10 million tonnes. With current operating capacity already at 6.20 million tonnes, Yunnan's hydropower carrying capacity is essentially at the ceiling — any new capacity must consume thermal, cross-province grid, or green-power certificates. Yunnan's "relocation window" is therefore narrowing.

Sichuan Ganzi's hydropower (lower Jinsha and Yalong basins) is abundant but at high altitudes within ecological zones, with development subject to dual oversight by MEE and Sichuan DRC. Current Ganzi capacity is only ~600,000 tonnes; the five-year growth ceiling is unlikely to exceed 400,000 tonnes. Qinghai Haixi and Ningxia Zhongwei PV projects, constrained by PV intermittency, must pair with storage (battery or pumped hydro) or thermal baseload — keeping per-tonne CO2 at 6-9 tonnes, well above pure hydropower's 2-3.5.

Inner Mongolia Tongliao and Ordos wind-PV plants face the "wind-PV-resource time-spatial mismatch" constraint. Wind output peaks in winter nights; PV output peaks in summer noon — neither matches the 24/7 baseload curve of electrolytic aluminum. Tongliao and Ordos wind-PV aluminum plants currently self-supply only 30-40 percent of demand, with the balance purchased from Mengdong Grid or captive coal-power. To lift wind-PV self-share above 80 percent requires large-scale storage (~100 MWh per 10,000-tonne capacity) or pumped hydro — adding RMB 50-80 million per 10,000 tonnes — raising wind-PV aluminum per-tonne capex about 20 percent above hydropower aluminum.

By international benchmarks, Norsk Hydro, Rio Tinto BC Works, Iceland Century, Brazil Albras, Oman Sohar, Mozambique Mozal, and Switzerland Alusuisse all run >95 percent green power with per-tonne CO2 at 2-4 tonnes — the international peer cohort for Yunnan Aluminium, Shenhuo Yunnan, and China Hongqiao Yunnan Wenshan Funing. While China's green-power aluminum capacity already leads in scale (Yunnan, Sichuan, and Qinghai combined ~7.20 million tonnes versus the Norway+Canada+Iceland combined 3.20 million tonnes), there's still a clear "soft-power" gap in green-power stability, ASI certification coverage, and green-aluminum brand recognition internationally.

The development of the green-power trading market is also vital infrastructure. State Grid and China Southern Power Grid have both set up provincial green-power trading platforms; electrolytic-aluminum firms can sign annual or multi-year PPAs directly with hydropower, wind, and PV producers in the province or across provinces. Yunnan's 2025 green-power trading volume was ~68 billion kWh, of which 45 billion kWh went to electrolytic-aluminum, supporting roughly 3.30 million tonnes of green-power-certified production. Green-power prices typically exceed common industrial tariffs by RMB 0.03-0.08 per kWh, but since green certification dovetails with ASI, Apple supply chain, and Tesla zero-carbon frameworks, the premium can be fully passed through to downstream. Yunnan Aluminium's per-tonne sales price includes a stable green premium of RMB 400-600.

H2 2025 also accelerated China's electricity-market reform — NDRC and NEA jointly issued the "Opinions on Deepening Market-based Electricity Trading," requiring electrolytic aluminum, calcium carbide, ferroalloy, chlor-alkali, and silicon to enter full marketization, with planned tariffs abolished. This reform's impact on coal-power-aluminum firms is enormous — Shandong, Henan, Ningxia, Xinjiang's previously enjoyed "base tariff + transmission" pricing is overturned; all power must be bid through electricity exchanges. Against the rapid rise of compute centers and data centers as new electricity-consuming entities, marketization may further lift the industrial tariff center, putting fresh pressure on coal-power-aluminum margins.

To grasp the actual landing of China's green-power replacement, see hydropower aluminum plants, green-power aluminum plants, and Yunnan electrolytic aluminum for the real factory map.

6. The recycled aluminum surge — scrap loops, low-carbon labeling, and EU CBAM

Recycled aluminum is the most-certain growth segment in China's aluminum value chain for the next decade, bar none. China produced 11.50 million tonnes of recycled aluminum in 2025, accounting for 22 percent of total aluminum supply; per MIIT's 2024-2027 Aluminum Industry High-Quality Development Plan, recycled-aluminum share is targeted to lift to 30 percent by 2030 — corresponding to 18-20 million tonnes annual recycled output.

The core advantages of recycled aluminum are energy use and carbon emissions. From scrap remelting to recycled aluminum ingot, unit electricity is ~5 percent of primary aluminum's; per-tonne CO2 is ~0.6 tonnes — a twenty-sixth of coal-power aluminum and a sixth of hydropower aluminum. From a global carbon-market perspective, recycled aluminum is the genuine "zero-carbon aluminum substitute." The IAI's Global Aluminium Cycle 2025 report notes global recycled aluminum output broke 32 million tonnes, 31 percent of total supply; EU is at 48 percent, US at 59 percent, Japan at 72 percent — all materially higher than China's 22 percent. This gap is the quantitative basis for the Chinese recycled-aluminum sector's next decade of growth.

China's recycled-aluminum raw-material sources fall into two buckets. "Old scrap" — from downstream-recovered scrap products, mainly used cans, construction scrap, appliance scrap, and automotive scrap — totaled 5.80 million tonnes in 2025. Used-can recovery rate is ~86 percent (above Japan's 93 but below US's 47); automotive scrap is entering its exponential release as NEVs age into scrap; construction scrap fluctuates with the property cycle. "New scrap" — trim, stamping waste, turning chips from processors themselves — totaled 5.70 million tonnes — its share scales linearly with processing output.

Concentration in China's recycled-aluminum sector is relatively diffuse — CR10 ~32 percent, far below electrolytic aluminum's 81 percent. Leading players include Ye Chiu Metal (601388.SH; Jiangsu Taicang plus Southeast Asia footprint; 1.20 million tonnes capacity), Sunbo Alloy (002996.SZ; Chongqing Tongliang; 700,000 tonnes alloy ingot), Xinjiang Joinworld (600888.SH; 600,000 tonnes deep processing), Mingtai Aluminium (battery-foil recycled-raw-material coordination), plus regional leaders like Suzhou Litai Precision, Anhui Yongjie New Materials, and Hilong Aluminium. Pain points: raw-material sources are scattered through fragmented scrap-recovery networks, with logistics atomization and inconsistent quality control, restricting downstream use to construction, casting, and other mid-low-end applications. High-end pathways are two: recycled aluminum alloy ingots feeding directly into automotive die-casting (which has already achieved high recycled-aluminum substitution), and closed-loop recycling of high-end sheet and plate (the "can-to-can" and "foil-to-foil" process loops).

From a deeper international comparison, EU's 48 percent recycled share rests on three pillars. First, the EU's scrap-recovery network covers the entire region, with consolidation high — Stena Recycling (Northern Europe), Remondis (Germany), and Veolia (France) together hold over 40 percent of EU scrap-metal recovery share. Second, EU's standardized "alloy-classified recovery" workflow keeps recycled-alloy composition stable and controllable, enabling direct use in automotive aluminum die-casting and other high-value applications. Third, the EU "Extended Producer Responsibility" (EPR) framework mandates that aluminum-product manufacturers assume a proportional recycling responsibility, driving "recyclability" into product design at source. China must close these three gaps simultaneously to catch up.

US's 59 percent share rests heavily on can recycling. The US consumes ~100 billion cans annually (310 per capita — globally the highest), with can recovery at ~47 percent; recovered cans flow through Constellium, Novelis, and others' "can-to-can" loops to be re-smelted into new cans. Japan's 72 percent share — the highest globally — reflects Japan's strict scrap-vehicle and scrap-appliance recovery regulations, with the world's highest recovery rates (automotive alloy at 96 percent, appliance alloy at 91 percent); Japan's recovered scrap goes mostly into automotive die-casting, where the closed loop is ~20 years more mature than China's.

China's three "bottleneck factors" are: fragmented scrap-recovery logistics (individual collectors dominate, scrap quality is inconsistent); weak alloy-classified recovery capability (most small recyclers still run "mixed smelting + alloy-blending" traditional processes); insufficient closed-loop agreements (no compulsory inter-firm recovery agreements between OEMs and aluminum companies; only Tesla, BMW, and Mercedes-Benz have piloted "old-for-new" alloy recovery in China). The next 5-10 years' core opportunity in Chinese recycled aluminum is the gradual breakthrough of these three bottlenecks — leading players (Ye Chiu Metal, Sunbo Alloy, Mingtai Aluminium) stand to gain meaningful share.

Battery-foil recovery remains exploratory. NEV battery cells contain ~2.5-3.5 percent aluminum foil by weight; at 5.20 million tonnes of 2025 China battery output, foil demand is ~1.40 million tonnes — of which ~720,000 tonnes is battery-grade high-purity foil. End-of-life battery-foil recovery faces the challenge of separating foil from cathode active material (LFP, NMC). CATL, BYD, Huayou Cobalt, Greenmei Recycle, and other top recyclers have rolled out multi-track approaches — physical disassembly, hydrometallurgical leaching, cryogenic fragmentation — with annualized scrap battery-foil recovery expected to reach 250,000 tonnes by 2028.

The high-value recycled aluminum used for automotive die-casting is the highlight segment. Currently ~40 percent of Chinese automotive die-casting raw material has been substituted with recycled alloy ingot — up from 18 percent in 2018, more than doubling in seven years. The core obstacle is composition stability — die-casting requires alloy element (Cu, Zn, Mg, Si, Fe, Mn) variation within 0.2 percent; otherwise mechanical properties and surface quality suffer. Ye Chiu Metal, Sunbo Alloy, Suzhou Litai Precision, and Anhui Yongjie New Materials have mastered "classified recovery + fine sorting + on-line spectrum detection + alloy blending + vacuum degas/dehydrogen treatment" to deliver high-end recycled ingot stably.

On low-carbon labeling, in 2024 Chalco, China Hongqiao, Yunnan Aluminium, Shenhuo, and Tianshan Aluminium together with CNIA launched the "China Green Aluminium" voluntary certification label, with third-party verification across four core indicators: green-power share, carbon intensity, recycled-aluminum share, and red-mud disposal. By year-end 2025, 19 electrolytic-aluminum plants, 12 alumina refineries, and 8 deep-processing plants had been certified — combined certified capacity around 9.80 million tonnes. ASI "Performance Standard" plus "Chain of Custody" double certification is the stricter international standard — since 2020, Apple has required all aluminum suppliers in its chain to obtain ASI certification; in 2025, Tesla, BMW, Daimler, Ford and others followed similar requirements. This international supply-chain pressure is the most-powerful external driver of China's "green-aluminum transition."

ASI Chain of Custody (CoC) detail is also worth unpacking. ASI CoC requires firms to trace aluminum products end-to-end from bauxite through electrolytic-aluminum to processing and downstream applications; every link's suppliers must hold ASI Performance Standard certification; delivered products carry "ASI qualification certificates" issued by independent third-party verifiers (BV, SGS, TÜV Rheinland, etc.). It places high demands on production records, raw-material origin, energy use, and carbon accounting data management. In China, only China Hongqiao, Nanshan Aluminum, and Yunnan Aluminium currently hold ASI CoC certification, and only for some sites. The slow extension of this international standard is the key bottleneck for Chinese aluminum firms in entering Apple/Tesla/BMW global low-carbon supply chains.

EU CBAM is the strongest external driver for the recycled-aluminum pathway. CBAM has imposed transitional reporting since October 2023 for aluminum products, and entered the formal-charging phase in January 2026. Charge basis: product "embedded emissions" valued at EU ETS spot price. At the 2025 EU ETS average of EUR 81 per tonne CO2, coal-power aluminum (16 tCO2 per tonne aluminum) faces an effective CBAM tariff of EUR 1,150 per tonne of aluminum; hydropower aluminum (3.5 tCO2) faces only EUR 250; recycled aluminum (0.6 tCO2) faces only EUR 40-50.

CBAM's 2026 landing has already shifted Chinese aluminum exports to EU — Q4 2025 China's EU-bound aluminum-product exports fell 8.4 percent year-on-year; Shandong and Xinjiang coal-power-aluminum producers fell 21 percent, while Yunnan and Sichuan green-power producers rose 12 percent. China's CBAM-response policy levers are three: accelerate national carbon-market expansion (formally including electrolytic aluminum from 2026); push for a "China-EU carbon-cost reciprocity mechanism" for partial mutual recognition; accelerate recycled-aluminum capacity to substitute primary-aluminum exports.

Notably, EU CBAM is not limited to primary aluminum — it gradually expands to processed and downstream products containing aluminum. The European Commission's February 2026 "CBAM Implementing Rules Amendment Draft" makes clear that CBAM will extend to aluminum extrusion, sheet, and foil from 2028, and to automotive and home-appliance finished products containing aluminum from 2030. This expansion materially affects Chinese auto and appliance exports — 2025 data shows Chinese auto exports embed ~680,000 tonnes of aluminum components, appliance exports embed ~1.40 million tonnes, totaling 2.08 million tonnes; on coal-power-aluminum basis, 2030 CBAM-equivalent additional tariff would be 12-18 percent of sale value.

Recycled-aluminum financing support also merits mention. China Development Bank, China Exim Bank, ICBC, ABC, CCB, BOC and other large lenders all run "green credit" dedicated quotas for recycled-aluminum capacity expansion, technology upgrade, and recovery-network buildout. By year-end 2025, green-credit cumulative lending to recycled aluminum totaled ~RMB 120 billion. Green-credit rates run 20-40 bps below standard commercial loans, materially easing recycled-aluminum financial burden. The CSRC, SSE, and SZSE also extend policy preference to recycled-aluminum IPOs and refinancing — qualifying "green low-carbon" firms enjoy IPO express channels and priority refinancing review.

Internationally, mature recycled-aluminum experience offers lessons. Alcoa Recycling is the world's largest recycled-alloy-ingot producer at 3.20 million tonnes annual capacity, covering North American automotive aluminum die-cast supply. Norsk Hydro's HyForge integrates Sapa extrusion process scrap and Nordic vehicle recovery into a "low-carbon aluminum alloy" closed loop. Japan's UACJ, Sumitomo Light Metal, and Mitsubishi Aluminium each run dedicated recycled-alloy ingot plants in Japan, serving Toyota, Honda, and Nissan. Germany's Trimet Aluminium, Austria's AMAG Metall, and France's Constellium have all built dedicated recycled-aluminum sub-segments. Their common lesson: recycled aluminum must form long-term closed-loop recovery agreements with downstream OEMs and battery makers to stabilize "can-to-can," "foil-to-foil," and "car-to-car" pathways.

International mutual recognition of "recycled-aluminum certification" is another emerging issue. EU "recycled-aluminum content label" requires traceable scrap origin and recovery-chain transparency. China's current "Recycled Metals Industry Access Conditions" specifies scale, energy use, and recovery rate, but not yet mandates origin traceability. This gap may be used by EU as a non-tariff barrier in coming years — imports of Chinese recycled-aluminum alloy ingot may need to carry "origin certificates," creating real barriers for Chinese recyclers without such certification. Chinese aluminum firms should pre-build "full-chain scrap-origin traceability" systems aligned with ASI and EU standards.

For the recycled-aluminum factory map, see recycled aluminum plants, scrap aluminum recovery, aluminum alloy ingot plants, and aluminum powder plants.

7. Our platform's view — searching real producing aluminum factories by process and downstream capability

The aluminum value chain is extraordinarily broad and deep — from open-pit mining in Boké (Guinea) to the battery-pack housing of a white electric vehicle in a Shanghai NEV dealership. Every node in the chain has many real operating factories — they may be alumina refineries near Boké port, electrolytic-aluminum plants in Wenshan (Yunnan), battery-foil slitting lines in Nantong (Jiangsu), or automotive aluminum die-casting plants in Foshan (Guangdong). In the Chinese market, identifying these real operating factories from business-registration data is itself a tough exercise.

Tianxia Gongchang is a B2B platform of 4.8 million real producing factories. The difference from generic business-registration services is this: those services can show you registered status but cannot tell you whether a given firm is actually a producing factory — many firms with "aluminum" in the name are construction-material traders, aluminum-product decoration outfits, or aluminum-alloy agents, not actually engaged in smelting or aluminum processing. Through a triple-verification mechanism (cross-validation across external data sources, AI-driven process-and-product classification, and on-site sample verification), our platform has identified 4.8 million real producing factories, classified into 1,965 sub-process/product categories, enabling precise filtering by process, capacity, or downstream application.

For the aluminum chain specifically, our platform supports three retrieval dimensions:

First, by process node. Alumina plants, electrolytic aluminum plants, prebaked anode plants, anode carbon plants, and recycled aluminum plants span various upstream and midstream nodes. Alumina refineries in China total about 30 with scale capacity — concentrated in Shandong (Weiqiao, Xinfa, TBEA), Shanxi (Chalco Xingxian, Huaxing Aluminium), Henan (Chalco Zhongzhou, Wanji Aluminium), Guangxi (Chalco Pingguo, Xinfa GuangTou), Guizhou (Chalco Guizhou, Guangxi Aluminium), and Yunnan (Yunnan Aluminium Funing, Hongqiao Wenshan) — easily enumerated through the platform.

Second, by downstream processing type. Aluminum sheet & plate plants, aluminum foil plants, aluminum profile plants, aluminum die-casting plants, aluminum billet plants, aluminum tube plants, and aluminum powder plants correspond to different downstream processor types. These factory counts vastly exceed upstream smelting — approximately 150 sheet-and-plate plants, 220 foil plants, 1,400 extrusion plants, and 2,800 die-casting plants — the long tail downstream of the aluminum value chain.

Third, by application domain. Automotive aluminum alloy, battery aluminum foil, aerospace aluminum, rail transit aluminum, can-stock aluminum, industrial aluminum extrusion, and aluminum composite panel plants cover diverse end markets. This dimension is particularly useful for the "reverse-engineer suppliers from end markets" research and commercial use case.

Combined with region and capacity filters — "Yunnan electrolytic-aluminum plants with annual capacity > 100,000 tonnes," "Jiangsu foil plants with battery-foil slitting capability," "Guangdong automotive aluminum die-casting plants" — these joint queries surface practically actionable target factory lists. This multi-dimensional structured retrieval is the differentiated value of our platform for the aluminum value chain.

Going deeper, our platform supports reverse retrieval by customer application — for example, "factories supplying battery housings for NIO ET5," "factories supplying PV frames for LONGi Green Energy," "factories supplying AC foil for Midea air conditioners." These reverse queries leverage our structured map of upstream-downstream transaction relationships — built from years of business-registration data, tax invoices, public tender records, product testing reports, customs trade flow, and brand-website disclosures. Such "downstream-to-upstream" reverse retrieval is uniquely valuable to industry research, banking due-diligence, investment fund due-diligence, government investment-attraction, and corporate supply-chain backup.

The real-factory identification system covers six dimensions: geographic (province, city, county); process (alumina refining, electrolytic smelting, casting, extrusion, rolling, foil rolling, annealing, anodic oxidation, etc.); capacity (micro, mid, large, super-large); product (alloy series — 1xxx pure aluminum, 2xxx Al-Cu, 3xxx Al-Mn, 4xxx Al-Si, 5xxx Al-Mg, 6xxx Al-Mg-Si, 7xxx Al-Zn, etc.); application (auto, battery, PV, rail, appliance, electronics, construction, packaging); enterprise type (listed, unlisted-private, foreign, state-owned).

By alloy series, our platform has built a complete product tree from primary alloy ingot to finished aluminum part. For example, 6063-series extrusion profile lookup returns ~1,100 supplier factories nationwide; 5052-series sheet/plate returns ~280; 3003-series foil returns ~190; 7075-series forging (for aerospace main structure) returns ~26; 1100-series high-purity aluminum (for capacitor and semiconductor sputter target) returns ~18. This "by-alloy" retrieval is invaluable for materials engineers seeking specific suppliers, researchers mapping supply chains, and procurement teams building multi-source backup.

Real-factory identification also addresses a key issue — "operating-state" real-time identification. A firm's business-registration status may show "live," but its actual operating state could be stopped, half-stopped, repositioned, or sub-let. Identifying real operating state requires cross-validating across dynamic data — tax invoice issuance (continuous-invoicing factories tend to be operating), customs records (continuous trade-flow factories tend to be operating), public tender records (continuous winning bids or supplier listings indicate operation), environmental supervision (continuous emission reporting indicates operation), electricity consumption (consumption matching designed capacity indicates operation). Our platform's identification rests on continuous monitoring and cross-validation of these five dynamic data streams.

Our coverage depth far exceeds traditional business registration services. For example, on "aluminum die-casting plant" node, beyond geography, capacity, customer-type filters, the platform supports finer-grained conditions like "factories with one-piece die-casting machines," "factories with 6000-ton+ presses," "factories with vacuum die-casting capability," "IATF 16949 certified factories." For "aluminum foil plant" node, finer filters include "battery-foil slitting line," "dual-layer coating foil capability," "8-micron ultrathin foil capability." This "by technical capability" granular retrieval is the key differentiator.

8. Mid-stream processing — sheet, foil, profile, automotive sheet, battery foil, can stock

Aluminum leaves the smelter as 8-25 kg ingots; to become car bodies, battery housings, beverage cans, PV frames, and rail vehicle profiles, it must pass through three major processing tracks: sheet-and-plate, foil, and extrusion. China's 2025 deep-processing output was 46 million tonnes — 16 million sheet-and-plate, 5.8 million foil, 24 million extrusion — combined, 82 percent of electrolytic-aluminum consumption.

Aluminum sheet-and-plate is delineated at 6 mm — above 6 mm is "plate," below is "sheet." Applications: automotive body (31 percent), consumer electronics (16 percent), appliance and kitchenware (19 percent), construction (12 percent), printing plate (8 percent), can stock (14 percent). Chinese sheet-and-plate CR5 is ~38 percent, with leaders including Nanshan Aluminum, Mingtai Aluminium, Novelis China, Shenhuo, and Chalco Luoyang Copper (sheet & plate business). Automotive sheet — the highest-value subsegment — has unit-area tonne price 2.2× that of construction sheet, plus 18-24 months OEM qualification cycles — meaningful entry barrier. Nanshan, Novelis China, and Mingtai together supply 76 percent of domestic automotive-sheet by leading OEMs.

Automotive-sheet core specs are 6xxx-series (Al-Mg-Si) for body outer panels (hood, roof, door outer), with good bake-hardening, anti-corrosion, and stampability. NEVs more often use 5xxx-series (Al-Mg) for battery pack housings and inner panels — 5xxx is more stampable but weaker on bake-hardening. A mainstream NEV (Model Y, ID.4, Song PLUS EV) uses ~95 kg of sheet — 14 kg roof, 22 kg doors, 16 kg hood, 33 kg battery-pack housing, 10 kg interior reinforcement. The 2025 automotive-sheet market is ~5 million tonnes, projected at 8.30 million tonnes by 2030 — CAGR 10.7 percent.

International competitive landscape: Novelis (Hindalco subsidiary) is the world's largest automotive sheet supplier at 1.80 million tonnes capacity, supplying GM, Ford, Tesla, BMW, Mercedes, Toyota, Honda, Hyundai, and nearly 100 OEMs. Constellium is global #2 at 900,000 tonnes, mainly serving Europe. Alcoa's Mill Products exited automotive sheet, with Arconic (renamed Howmet post-spin-off) continuing operations. Japan's Kobe Steel, UACJ, and Sumitomo Light Metal are Japan's three suppliers. China's local players led by Nanshan, Novelis China, and Mingtai have achieved ~76 percent localization — premium-segment (NIO ET9, Li L9, AITO M9, Xiaomi SU7) body aluminum sheet still partly relies on Novelis imports.

The aerospace alloy sheet localization journey deserves attention. Aerospace alloys (2024, 2124, 7050, 7150, 7055 series high-strength alloys) are core materials for aircraft wings, fuselage skins, fuselage frames, and structural parts, with stringent strength, fatigue resistance, fracture toughness, and forming requirements. Aircraft main-structure alloy sheet had been almost entirely supplied by Alcoa, Constellium, and Rio Tinto's Canadian Alouette. Nanshan launched its Yantai aerospace alloy large-thickness project in 2018, achieved volume production of 7050-series in 2021, and industrial-scale production of 7150 and 7055-series in 2023; by 2025, it had become a core supplier to COMAC C919 and ARJ21 aerospace structural materials and supplies AMS-certified high-end aerospace sheet to Airbus Tianjin assembly and Boeing China aftermarket. Nanshan's aerospace sheet capacity is 65,000 tonnes — far below Alcoa's global aerospace capacity of ~400,000 tonnes — but domestic share now exceeds 60 percent.

Aluminum foil — delineated at 200 microns: below 200 microns is "foil," above is "strip." 2025 Chinese foil output 5.80 million tonnes — battery foil 720,000 (+38 percent YoY), pharma foil 250,000, household foil 220,000, packaging foil 1.90 million, capacitor foil 80,000, AC foil 600,000, others 2.03 million. Battery foil is the fastest-growing — 320,000 tonnes in 2022 to 720,000 in 2025, CAGR 31 percent. Suppliers include Dingsheng (603876.SH — global #1), Shenhuo, Mingtai, Huafon, Asia-Pacific, Wanshun, Dongyangguang Aluminium (top-three globally in capacitor foil).

Battery foil spec is 10-15 microns; thinnest at 8 microns. Production requires roll-surface precision <0.1 microns and shape symmetry <0.3 percent. The 2025 emergence of "dual-layer coated foil" — applying carbon-based or polymer coating to foil surface to suppress electrolyte corrosion of cathode current collectors — extends battery cycle life by 12-18 percent. Dingsheng, Huafon, and Wanshun have all small-batch produced dual-layer coated battery foil. By 2028, Chinese battery-foil demand should break 1.20 million tonnes.

Battery-foil technology barriers also lie in production equipment. The core foil-mill equipment is concentrated among three international suppliers — Hitachi Metals (Japan), Achenbach Buschhütten (Germany), and ANDRITZ (Austria). China's Chalco Luoyang Copper, Taiyuan Stainless Steel Machinery, and Erzhong Heavy Machinery have broken through six-roll precision finishing mill localization, but 12-roll high-precision foil slitting lines remain import-dependent. A single Hitachi 12-roll mill costs ~RMB 120 million — the single-largest CapEx item for battery-foil firms.

Foil slitting lines are critical battery-foil equipment. Battery-foil products must be slit per battery-maker's specific dimensions (prismatic 100×200 mm, cylindrical 21 mm diameter, etc.) with slitting precision <0.3 percent shape symmetry and burr height <5 microns. Slitting-line core is high-precision slitting blades and tension control, supplied by Mitsubishi Heavy Industries (Japan), Pittsburgh (Germany), and ANDRITZ (Austria). An 8,000-tonne-year battery-foil slitting line costs ~RMB 50 million — key CapEx for battery-foil firms.

Aluminum extrusion — pressing alloy billet through dies — is the largest deep-processing subsegment by output. 2025 Chinese extrusion 24 million tonnes — building 14 million (windows, curtain walls, decoration), industrial 8 million (rail vehicles, NEV structures, PV frames, machine frames, appliance components), other 2 million. PV frame is industrial-extrusion's fastest-growing — 1.10 million tonnes in 2021 to 3.20 million in 2025, projected at 5 million by 2030. PV-frame leaders include Xingfa Aluminium, Asia-Pacific, Fortune-Ness Holdings, Yongzhen Technology, Mindong Aluminium. Rail-transit extrusion is dominated by CIA-Sapa (Chalco joint venture with Sweden's Sapa), Zhongwang Holdings (post-restructuring), and Guangdong He-Sheng Industrial Aluminium — main suppliers to CRRC's high-speed sets, subway cars, and trams.

PV-frame technology upgrade is also worth tracking. Traditional PV frames use 6063 alloy; from 2025, upgrade to 6082 and 6005A higher-strength alloys aligns with bifacial PV (front-and-rear glass without backsheet) lightweighting requirements. Another upgrade direction is surface oxidation and corrosion resistance — 25-year module life requires 8,000+ hour salt-spray corrosion life for frames, far above the 2,000 hours typical of construction windows. Yongzhen Technology, Fortune-Ness Holdings, and Xingfa Aluminium have broken through 8,500-9,000 hour salt-spray life, supplying LONGi, JA Solar, JinkoSolar, Trina, Canadian Solar.

Aluminum profile sheet-and-plate width specs deserve a deeper look. From thickness, sheet-and-plate spans heavy plate (6 mm+), middle plate (2-6 mm), thin sheet (0.5-2 mm), and ultra-thin sheet (0.2-0.5 mm). Heavy plate is for construction curtain walls, structural parts, rail vehicles; middle plate for automotive body, electronics, appliances; thin sheet for can stock, printing plates, plastic-aluminum packaging; ultra-thin already abuts foil (below 200 microns), used in high-end packaging. By alloy series — 1xxx (pure aluminum) for can-bottom closure, electrical busbar, chemical vessels; 3xxx (Al-Mn) for can body, AC foil, household foil; 5xxx (Al-Mg) for body inner panels, battery housings, marine platforms; 6xxx (Al-Mg-Si) for body outer panels, construction profile, automotive structures; 7xxx (Al-Zn) for aerospace main-structure; 2xxx (Al-Cu) for aerospace interior, military aluminum alloy.

Chinese sheet-and-plate sector equipment configuration shows clear "tier differentiation." Leaders (Nanshan, Mingtai, Novelis China, Shenhuo, Huafon) deploy SMS, ANDRITZ, etc. top-tier international suppliers' latest four-roll or six-roll mills, with shape precision <3/1000; mid-tier (Asia-Pacific, Wanshun, Dongyangguang Aluminium, Xinjiang Joinworld, Hilong Aluminium) deploy mid-tier international or Chinese leading suppliers (Chalco Luoyang Copper, Taiyuan Heavy Industries), with precision 5-8/1000; smaller firms (~1,000 across) deploy domestic generic or refurbished mills, with precision above 10/1000. This equipment-tier difference determines product-quality distribution — leaders monopolize high-end automotive sheet, aerospace plate, battery foil; smaller firms cluster in construction, appliance.

Can stock (3xxx body + 5xxx top/tab + 1xxx bottom closure multi-alloy system) is the high-purity bulk of sheet-and-plate. 2025 Chinese can-stock 2.20 million tonnes — mostly for beer, cola, energy drinks, coffee drinks. A 330ml can weighs ~13.5 g — body 9.8g, top 3.2g, tab 0.5g. Can-stock recovery rate is the world's highest (China 86 percent, Brazil 94 percent), the most-reliable recycled-aluminum raw material. Mingtai, Novelis China, and Shenhuo are China's three main can-stock suppliers.

Rail-transit aluminum is the representative of high-end industrial extrusion. CRRC's 2025 new high-speed sets (380 8-car), subway cars (5,120), trams (320) consumed ~260,000 tonnes alloy extrusion. Rail alloy is mainly 6005A and 6082 extrusion profile, requiring high strength, fatigue resistance, weldability. CIA-Sapa (Chalco joint venture with Hydro Sapa) is the largest rail-transit alloy supplier in China at 120,000 tonnes annual capacity, supplying CRRC's Changchun Railway Vehicles, Qingdao Sifang Locomotive, Tangshan Railway Vehicles, Zhuzhou Electric Locomotive, and Nanjing Puzhen Vehicles.

Rail-transit application scenarios deserve unpacking. A CRH high-speed set (8-car) uses ~22 tonnes alloy extrusion. A subway set (6-car) uses ~18 tonnes. A tram set (3-car) uses ~9 tonnes. Tech barriers: cross-section complexity (some sections include 10+ cavities and reinforcement ribs), weldability (friction-stir-welding plus MIG/MAG requirements), fatigue (30-year design life, millions of cycles), corrosion resistance (tunnel humidity and salt spray).

Automotive die-casting is a high-volume deep-processing subsegment. 2025 Chinese automotive alloy die-casting was ~6.50 million tonnes total — auto applications 62 percent (4.03 million), including engine block, transmission housing, chassis suspension, one-piece body cast, battery tray, motor housing. CR10 is ~36 percent; leaders include Hongtu Technology (002101.SZ), Wencan Group (603348.SH), Aikedi (600933.SH), Xusheng Group (603305.SH), Qunfeng Auto (603982.SH), Lizhong Group (300428.SZ), Rongtai Shares (605133.SH), Ruihu Mould (002997.SZ).

One-piece die-casting is the emerging sub-trend. Since Tesla pioneered with the Model Y rear underbody (8,000-ton press) in 2021, NIO ET5, XPeng G6, AITO M9, Xiaomi SU7, Li L9 have all followed. A one-piece part weighs 60-90 kg, using heat-treatment-free alloy (Tesla's proprietary Tesleon, Lizhong LSAC-365, Wencan WC-360 etc.), replacing 60-90 traditional stamped-welded parts. Tech barriers: heat-treatment-free alloy formulations and super-large press process control. Leaders: Wencan (603348.SH), Hongtu Technology, Lizhong, Aikedi.

Aluminum extrusion anodic oxidation is the critical post-treatment, with decisive impact on appearance, anti-corrosion, and wear resistance. Anodizing is an electrochemical process — passing current with the alloy as anode in sulfuric or chromic electrolyte — to grow a dense Al2O3 oxide film on the surface. Film thickness usually 10-25 microns, lifting salt-spray corrosion life from generic alloy's 200 hours to 10,000+ hours. Leaders: Guangdong Yaluminum (construction profile anodizing), Fujian Mindong Aluminium (industrial extrusion anodizing), Inner Mongolia Fengchuan (PV frame anodizing), Henan Sanmenxia Qiaozhi (appliance alloy anodizing). Anodizing wastewater (aluminum, sulfuric, chromic) management is a key environmental challenge — RMB 6-8 per tonne water treatment cost, lower than typical electroplating (RMB 10-12) but mismanagement causes water pollution.

Home-appliance alloy and foil markets also shift structurally. China is the world's largest producer of AC, refrigerators, washing machines, water heaters — 2025 home-appliance aluminum consumption ~6.80 million tonnes: AC foil (condenser, evaporator fins) 1.20 million, refrigerator foil 250,000, washing-machine alloy 400,000, water-heater alloy 280,000, other 4.65 million. Major suppliers: Gree's Kaibang Motor (captive), Midea's Meizhuo Aluminium (captive), Haier's Rixinshun (captive), Huafon (independent third-party). Sub-trend: ultra-thinning — AC foil from 0.10 mm in the 1990s to 0.05 mm today, further lightweighting downstream appliance weight and alloy use.

Aluminum billet (round) is also worth a dedicated mention. Billet is the core feedstock for extrusion — diameter 85-300 mm, length 3-6 m, alloys mainly 6063, 6005A, 6082, 6061, 7075. 2025 Chinese billet output ~26 million tonnes — 6063 for construction 14 million, 6005A for PV frame 3.50 million, 6082 for automotive structures 2.80 million, others 5.70 million. Billet processing — DC casting, ingot heating, extrusion (on 4,000-12,000 ton presses), annealing, surface treatment. Leaders: Xingfa Aluminium (construction billet), Asia-Pacific (industrial), Huafon (automotive), Yongzhen (PV frame billet), Mindong Aluminium (comprehensive).

To observe China's deep-processing supply-chain map, see automotive sheet, extrusion plants, PV frame aluminum, aluminum single-panel plants, and AC foil for the breakdown by region and capacity.

9. Capacity expansion — Yunnan Aluminium hydropower, Tianshan Fukang, Shenhuo Yunnan dismantling

2025-2026 is the critical transition window for Chinese electrolytic-aluminum capacity. Under the 45-million-tonne ceiling, "relocation" is the only viable expansion path. Three landmark events define the period: Yunnan Aluminium's Zhaotong Phase 2 commissioning, Tianshan Aluminium's Xinjiang Fukang Phase 1 commissioning, and Shenhuo Shangqiu Phase 2 dismantling-and-relocation to Yunnan Zhaotong.

Yunnan Aluminium's Zhaotong Phase 2 — at Yongshan County, Zhaotong — draws on Jinsha River Baihetan dam and Yongshan Huangyangpo Hydropower. Designed at 500,000 tonnes/year. Phase 1 (500,000 tonnes) was commissioned in August 2025; Phase 2 (500,000 tonnes) in January 2026. Hydropower share 92 percent, DC unit power 12,800 kWh/tonne, per-tonne CO2 only 3.2 tonnes — the lowest in China. Post-Phase-2, Yunnan Aluminium total capacity reached 3.30 million tonnes, ranking 4th in China (after Chalco, China Hongqiao, and Chalco Qinghai Aluminium under China Electric Power Investment).

Tianshan Aluminium's Xinjiang Fukang Phase 1 — at Fukang, Changji Hui Autonomous Prefecture — draws on Junggar coal-power base and captive coal-power. Designed at 600,000 tonnes/year. Commissioned in September 2025. Importantly, Fukang is not "new capacity" — Tianshan acquired the indicator from a Qinghai exit via indicator swap. Per-10,000-tonne indicator deal price ~RMB 170 million — reflecting the discount-present-value of per-tonne profit. Fukang's combined power cost is just RMB 2,700 per tonne, among China's lowest.

Shenhuo's Shangqiu Phase 2 dismantling-and-relocation to Yunnan Zhaotong is a landmark in Chinese cross-province capacity relocation. Shenhuo originally ran 850,000 tonnes at Shangqiu (Henan), with combined cost of ~RMB 16,800 and per-tonne profit of ~RMB 1,800. From 2020-2022, Phase 1 relocated 450,000 tonnes to Yunnan Shenhuo, capturing low Yunnan hydropower tariff (RMB 0.35/kWh, 32 percent below Henan's RMB 0.52) and green premium (RMB 300-500/tonne). 2025-2026 sees Phase 2 — dismantling and relocating the remaining 400,000 tonnes from Shangqiu to Zhaotong. Post-relocation, Yunnan Shenhuo will reach 1.30 million tonnes and Shangqiu will go to zero.

Other notable 2025-2026 capacity moves: East Hope's Inner Mongolia Tongliao Huomei expanded from 500,000 to 700,000 tonnes, with 200,000-tonne indicator from Qinghai West Mining's exit at ~RMB 180 million per 10,000 tonnes. China Hongqiao's Yunnan Wenshan Funing Phase 4 (500,000 tonnes) commissioned in March 2026 — completing the fourth phase of Shandong-to-Yunnan "million-tonne cross-province relocation." Xinfa Group's Shandong Liaocheng smelter started "coal-to-gas + green-power purchase" upgrade, cutting per-tonne CO2 from 16 to 9.8 tonnes and partially capturing EU green-aluminum demand. East Hope's Xinjiang Changji Ji Phase 2 (600,000 tonnes) commissioned in Q4 2025, lifting Xinjiang capacity to 6.50 million tonnes — overtaking Shandong as China's second-largest electrolytic-aluminum province.

Geographic center-of-gravity shift: 2025 China's "four major regions" — Yunnan (6.20 million), Xinjiang (6.50 million), Inner Mongolia (6.80 million), Shandong (4.40 million). Versus 2017's "Shandong dominant (8.90 million) + Xinjiang (3.80) + Shanxi (5.40) + Henan (3.20)" — entirely different. Green-resource provinces (Yunnan, Sichuan, Qinghai) combined electrolytic-aluminum capacity broke 8.50 million tonnes, 19.2 percent of national total, up 12 points from 7 percent in 2017. By 2028, green-power aluminum share will likely climb to 25-30 percent.

The capacity-expansion CapEx intensity merits dedicated calculation. A 500,000-tonne electrolytic-aluminum plant requires CapEx of ~RMB 12-15 billion: cell construction RMB 5-6 billion, casting shop RMB 1-1.2 billion, melt transport RMB 200-300 million, power and substation RMB 2-2.5 billion, auxiliaries and structures RMB 3-4 billion. Captive coal-power (Xinjiang, Ningxia) adds another RMB 8-10 billion; hydropower aluminum projects typically secure hydropower via long-term PPA without building their own plants. Yunnan Aluminium Zhaotong Phase 2 (500,000 tonnes) had CapEx of RMB 13 billion (RMB 26,000 per tonne); Tianshan Aluminium Xinjiang Fukang Phase 1 (600,000 tonnes with captive coal-power) had CapEx of RMB 22 billion (RMB 36,700 per tonne); China Hongqiao Yunnan Wenshan Funing Phase 4 (500,000 tonnes) had CapEx of RMB 12 billion (RMB 24,000 per tonne).

Domestic deep-processing capacity dynamics are also active in 2025-2026. New capacity: battery foil, automotive sheet, PV frame, automotive die-casting cluster expansion — 2025 added ~280,000 tonnes battery foil, ~170,000 tonnes automotive sheet, ~800,000 tonnes PV frame, ~260,000 tonnes die-casting. Exit capacity: construction extrusion, appliance sheet, low-end foil — 2025 exited ~1.10 million tonnes extrusion, ~300,000 tonnes appliance sheet, ~200,000 tonnes low-end packaging foil. This "high-end in, low-end out" capacity reshuffle is the next-five-year mainline of China's aluminum value chain.

Overseas capacity expansion: 2026-2030 overseas projects deployed: Indonesia — Tianshan Aluminium Bintan Phase 2 alumina (1 million tonnes, H2 2026), China Hongqiao Galang Batang Phase 3 alumina (1 million tonnes, 2027), East Hope Sulawesi electrolytic-aluminum (1.50 million tonnes, 2028); Guinea — Chalco Boffa alumina (1.20 million tonnes, 2027), China Hongqiao SMB alumina (1 million tonnes, 2029); Iran — China Steel International's Iralco Phase 2 electrolytic (600,000 tonnes, 2028); Turkey — Nanshan Aluminum Kayseri electrolytic (400,000 tonnes, 2029); Mexico — Dingsheng Monterrey battery foil (80,000 tonnes, 2027); Hungary — Ningbo Xusheng's Sino-Innovation Aluminium die-casting (120,000 tonnes, Q4 2026), Aikedi Budapest die-casting (80,000 tonnes, 2027).

International capital pathways are diversified. Chinese aluminum firms' overseas-project funding sources include: organic cash flow, A-share or H-share secondary offerings, banking M&A loans, Silk Road Fund, AIIB loans, commercial-bank long-term loans, and offshore bond issuance. Silk Road Fund and AIIB have provided particularly strong support for "Belt and Road" projects in Guinea, Indonesia, and Iran — Silk Road Fund has equity-invested over USD 3 billion across China Hongqiao Guinea bauxite, Chalco Boffa alumina, and East Hope Indonesia Sulawesi electrolytic-aluminum; AIIB has lent over USD 2 billion across Indonesia Galang Batang alumina, China Hongqiao SMB bauxite, and other projects.

Construction cycle: a 500,000-tonne electrolytic-aluminum plant takes 24-30 months from land to commissioning — 6 months preliminaries, 10 months civil and equipment ordering, 6 months installation and commissioning, 8 months trial and ramp-up. Main bottleneck: not engineering capability (Chalco, China Hongqiao, East Hope, and Shenhuo have built dozens of plants between them), but indicator approval, land acquisition, and power/substation connection. Yunnan Aluminium Zhaotong Phase 2 took ~26 months; Tianshan Fukang Phase 1 ~28 months (Xinjiang grid expansion delayed); Shenhuo Yunnan Phase 2 dismantling ~34 months due to Shangqiu phased decommissioning.

Indicator trading market detail: three typical modes — direct firm-to-firm negotiation (price reflects seller exit motive and buyer demand intensity), local-government-led "regional swap" (within-province only; Henan, Shanxi, Guizhou etc.), and the new "China Electrolytic Aluminum Capacity Indicator Trading Center" (Kunming, founded by Chalco, China Hongqiao, Yunnan Aluminium etc., providing standardized indicator trading). 2025 Center cumulative volume ~520,000 tonnes at ~RMB 90 billion — the industry's price benchmark.

2026 capacity ceiling enters "substantively maxed" stage, market dynamics likely to shift. H2 2026 to 2027: NDRC and MIIT may jointly release ceiling-management refinement — e.g., "loose ceiling" for green-power capacity (allowing hydropower/wind-PV projects to bypass single-firm cap), "strict ceiling with accelerated exit" for coal-power capacity. This refinement would further accelerate "coal-power exit + green-power expansion" capacity reallocation. Indicator markets may then bifurcate — single-10,000-tonne coal-power indicators may fall below RMB 100 million, while green-power indicators may climb to RMB 250-300 million.

To grasp China's electrolytic-aluminum capacity expansion landing, see Yunnan aluminum plants, Xinjiang aluminum plants, Inner Mongolia aluminum plants, and Shandong aluminum plants for the breakdown.

10. Price cycle — 2024-2026 LME, SHFE, alumina, and power

Pricing is the core driver of Chinese aluminum profits. Understanding 2026 aluminum-price direction requires a full 2024-2026 cycle view across LME 3-month, SHFE front-month, alumina spot, bauxite CIF, and key energy costs.

LME 3-month aluminum averaged USD 2,420/tonne in 2024, up 7.1 percent from 2023's USD 2,260. The 2024 trading range was USD 2,080-2,820, full-year amplitude 35.6 percent. 2025 LME entered an uptrend at USD 2,690 average, up 11.2 percent. Q1 2026 saw resonance among AI compute demand, CBAM, Guinea export policy, and Yunnan drought expectations — March 2026 broke above USD 2,950/tonne, the highest since May 2011. Q2 2026 saw retracement to USD 2,810 as Yunnan rainy-season recovery, Guinea policy clarification, and global tight-money pressure on downstream consumption took hold.

SHFE front-month aluminum tracks LME tightly, but exchange-rate, inventory seasonality, and VAT-deduction chain effects yield independent "SHFE-LME ratio" movement. 2024 SHFE average RMB 18,500; 2025 RMB 19,600 (+5.9 percent); Q1 2026 peak RMB 21,800; Q2 2026 retraced to RMB 20,900. The Q1 2026 ratio briefly fell to 7.2 (SHFE lower than LME), reflecting LME's strength on international compute demand; by Q2, ratio recovered to 7.5, back to historical central tendency.

Alumina is electrolytic aluminum's largest input — 1.95 tonnes alumina per tonne electrolytic aluminum. 2025 alumina spot average RMB 4,150/tonne, but internal volatility was extreme — low RMB 3,180 (March 2025), peak RMB 5,650 (December 2025), amplitude 77 percent. Driver: Guinea export policy adjustments, China new alumina capacity ramp timing, smelter inventory cycle. Q1 2026 saw Chalco Zhongzhou Henan 3 million tonnes, Huasheng Materials Guangxi 2 million tonnes, and Tianshan Bintan Indonesia 2 million tonnes — alumina supply improved, price retraced to RMB 3,950.

Bauxite CIF China reflects Guinea-to-China shipping-chain tightness. 2024 Guinea CIF average USD 68; 2025 USD 76; January 2026 peak ~USD 92; by June 2026 retraced to USD 78. Australia CIF China 2025 USD 82 — slightly higher than Guinea but shorter shipping.

Power cost is electrolytic aluminum's largest, most-sensitive variable cost. Industry average DC consumption is 13,500-14,500 kWh/tonne; with industrial tariffs at RMB 0.35-0.55/kWh, power is 35-45 percent of combined cost. Specifically: Yunnan hydropower aluminum at RMB 0.30-0.35/kWh (including transmission); Xinjiang coal-power at RMB 0.22-0.25 (captive); Shandong coal-power at RMB 0.45-0.55 (purchased from State Grid Southern); Inner Mongolia wind-PV at RMB 0.32-0.40 (captive wind-PV + grid). The regional tariff gap converts directly to per-tonne aluminum cost gaps of RMB 2,000-3,000.

Prebaked anode price is another 2025-2026 cost variable. 2025 average RMB 5,400/tonne — per-tonne aluminum anode consumption cost RMB 2,200; 2026 with petroleum coke rising (Saudi Aramco coke export quota adjustment, Chinese refinery coke production decline), anode hit RMB 5,800 — per-tonne cost up RMB 160. Sunstone Development, Shandong Qili, and Guangxi Qiangqiang Carbon all saw Q1 2026 GMs widen ~2.8 ppt.

LME aluminum's seasonal pattern is also worth mentioning. Historically: Jan-Mar — China's post-Spring-Festival downstream recovery drives gradual upward bias; Apr-Jun — construction-and-appliance peak season holds firm; Jul-Aug — summer heat and downstream off-season pull back; Sep-Oct — "Golden Sept and Silver Oct" peak rebounds; Nov-Jan — winter off-season and year-end stock pressure pull back. The 2021-2023 "compute revolution" disrupted this — data-center construction's rapid demand growth turned Q3-Q4 into "abnormally strong" seasons. 2026's seasonal pattern is expected to deviate further from historical means, given multi-variable interactions among compute, CBAM, NEV, and others; pure seasonal-arbitrage strategies face higher risk.

Alumina-aluminum price "decoupling" is another notable financial phenomenon. Theoretically, alumina and aluminum are upstream-downstream — alumina rising should drive aluminum higher. In practice, alumina-aluminum correlation is only ~0.58, well below perfect. Reason: alumina is driven mainly by bauxite supply and alumina capacity timing; aluminum is driven by global macro and downstream demand — different drivers. December 2025 saw alumina rise 52 percent on Guinea export policy and Chinese alumina ramp mismatch; aluminum rose only 12 percent — compressing smelter alumina-aluminum margin sharply.

Global price variables also include US dollar index, oil price, freight, and electricity-market pricing. Dollar strength typically compresses LME aluminum (USD-denominated); oil price rises lift petroleum coke and tar pitch, raising prebaked-anode cost; freight rises lift Guinea bauxite CIF; electricity-market pricing directly affects coal-power-aluminum tariff base. 2026 reads "high dollar, mid oil, volatile freight, rising power" — a mild support for aluminum prices.

Major investment banks' H2 2026 to 2027 views: Morgan Stanley's June 2026 Global Aluminium Outlook H2 2026 maintains "neutral-positive" — LME H2 2026 USD 2,700-2,900, 2027 USD 2,600-2,800. Goldman Sachs' May 2026 Commodity Watch Aluminium is more bullish — LME H2 2026 above USD 3,000, 2027 mid-USD 2,900. JPMorgan is more neutral — H2 2026 USD 2,650-2,850. HSBC, UBS, Barclays are neutral-conservative. Combined, LME H2 2026 to 2027 center is likely RMB-equivalent USD 2,700-2,900, with downside risk from European/N. American downstream weakness, upside from Yunnan drought and Guinea export policy.

The financial-real-economy interaction is also notable. Chinese aluminum firms' cash flow, debt structure, profit distribution, M&A — these "financial decisions" are highly correlated with aluminum price cycles. In upswings (2021-2022, 2025-Q1 2026), leading firms see ample cash flow, falling debt, rising payout ratios, accelerating overseas M&A; in downswings, cash flow tightens, debt rises, payout falls, overseas expansion pauses. This couples Chinese aluminum value chain's medium-term path with global liquidity, FX, and credit cycles. Understanding Chinese aluminum's medium-term investment value requires analyzing both aluminum fundamentals and financial-cycle variables.

2025-2026 also saw shifts in "SHFE aluminum position structure" worth tracking. As of June 2026, SHFE aluminum main-contract open interest is ~620,000 lots (3.10 million tonnes), with speculative positions (CTA, macro hedge funds, retail) at ~42 percent and hedging positions (smelters, processors, traders) at ~58 percent. Speculative share rose 7 ppt from 2021's 35 percent — reflecting rising financialization. This shift heightens short-term sensitivity to macro mood, dollar cycle, geopolitical events; firms' basis risk in hedging is correspondingly amplified.

Aluminum option market also developed rapidly in 2023-2025. SHFE aluminum options launched in August 2020; 2025 daily average volume ~200,000 lots (1 million tonnes) — an important risk-management tool. Implied volatility averaged 21 percent in 2025; Q1 2026 briefly hit 32 percent on price expectations; Q2 2026 retraced to 24 percent. Option volatility is another sentiment indicator — significantly above historical mean often signals market disagreement or major uncertainty.

Total: 2025 Chinese electrolytic-aluminum industry average per-tonne net profit ~RMB 1,800; Yunnan hydropower aluminum ~RMB 2,400; Xinjiang coal-power ~RMB 2,100; Shandong coal-power ~RMB 900. Q1 2026 price upcycle saw industry average climb to ~RMB 2,500; Yunnan hydropower peaked near ~RMB 3,200 — the fundamental economic driver behind Yunnan Aluminium, China Hongqiao Yunnan Wenshan Funing Phase 4, and Shenhuo Yunnan Phase 2 ramp.

Price-cycle transmission to deep processing shows "short squeeze, long pass-through." High-value processed products (automotive sheet, battery foil, can stock) — processing fee is ~35-40 percent of sale price — low sensitivity to aluminum-price fluctuation, GM stable at 18-24 percent. Mid-low-value processed (construction extrusion, appliance sheet) — processing fee only 12-18 percent — GM only 6-9 percent — highly sensitive. See aluminum processing plants, aluminum alloy billet plants, high-purity aluminum plants.

11. Policy — dual carbon, CBAM, national carbon market, 45-million-tonne ceiling

Chinese electrolytic aluminum is policy-dense — capacity ceiling, dual-carbon goals, national carbon market, CBAM response, green-aluminum certification, recycled-aluminum policy together form the five-pillar policy ensemble for 2025-2026.

The 45-million-tonne ceiling stems from the 2017 NDRC-MIIT joint Notice on Accelerating Supply-Side Structural Reform and Effectively Resolving Excess Capacity in Electrolytic Aluminum, which set four principles: no new capacity, equivalent swaps for replacement, eliminate obsolete capacity, coal-power-aluminum coupling. By year-end 2025, operating capacity 44.20 million tonnes, total built capacity 44.56 million — leaving only ~440,000 tonnes of headroom. The 2026 ramp of Yunnan Aluminium Zhaotong Phase 2, Shenhuo Yunnan Phase 2, East Hope Changji Phase 2, and China Hongqiao Wenshan Funing Phase 4 pushes the system into "substantively maxed" mode. H2 2026 will likely see NDRC start a capacity-ceiling adjustment study, but 2026-2028 likely maintains the 45-million-tonne ceiling.

On dual-carbon, the 2020 Xi Jinping announcement of "carbon peak by 2030, neutrality by 2060" was followed by MIIT/NDRC joint Industrial Sector Carbon Peak Implementation Plan and Nonferrous Metals Industry Carbon Peak Implementation Plan, setting per-tonne energy intensity reduction targets of 5 percent by 2025 vs 2020, peak then decline by 2030, sectoral neutrality by 2060. MIIT's 2024 Aluminum Industry High-Quality Development Plan (2024-2027) sets three quantitative targets — green-aluminum share, recycled-aluminum share, ultra-low-emission plant count — at 30 percent, 25 percent, and 20 plants by 2027.

National carbon market (CEA) expansion is the most important 2026 policy variable. By year-end 2025, CEA covered only the power sector, ~4 billion tonnes annual emissions. From early 2026, electrolytic aluminum — a "high-energy, high-emission, high-mitigation-potential" sector — was named first-batch expansion target. Year-1 (2026): free quota at 95 percent; firms only purchase 5 percent; covered emissions ~700 million tonnes. Year-2 (2027): free quota down to 90 percent. Free quota declines yearly to 75 percent by 2030. At current CEA price RMB 90/tonne CO2, 2026 sector compliance cost ~RMB 3 billion (RMB 70 per tonne aluminum). By 2030, that cost climbs to RMB 300-500 per tonne aluminum.

CBAM response: EU CBAM transitional reporting from October 2023; formal-charge from January 2026, charged per embedded CO2 at EU ETS spot. 2025 EU ETS average EUR 81; Q1 2026 EUR 92. Chinese 2025 aluminum-product exports to EU were 610,000 tonnes — mostly foil, alloy sheet, and processed products; coal-power-aluminum-origin exports face CBAM equivalent to EUR 1,100-1,200 per tonne aluminum — losing EU price competitiveness. NDRC/MIIT/MEE jointly launched a CBAM-response working group, with three priorities: build a "China-EU carbon cost reciprocity mechanism" for partial mutual recognition; accelerate national carbon-market expansion and price ascension to close the gap with EU ETS; encourage high-carbon-aluminum relocation to Yunnan hydropower or green-power purchase upgrade.

International benchmarking: EU EU ETS is the most-mature global carbon market, covering all high-emission sectors, price stable at EUR 80-100. US has no nationwide carbon market but California Cap-and-Trade and Northeast RGGI run >10 years at USD 30-40 per tonne. 2025 Biden Clean Competition Act proposed a "carbon adjustment fee" on high-carbon industrial imports but not passed by Congress. Canada applied national carbon pricing from 2022 at CAD 40-65 per tonne, and from 2026 levies additional fees on above-benchmark industrial emissions. India hasn't yet built a national carbon market, but energy-efficiency certificates (PAT scheme) include aluminum. Overall, EU ETS + CBAM is the strictest and most-spillover global carbon-pricing system — Chinese aluminum internationalization must address CBAM as its first agenda.

Chinese carbon market "quota allocation method" deserves unpacking. CEA quotas use "industry-baseline method" — based on industry historical average carbon intensity, firms above baseline buy quotas, below baseline sell. Relatively fair in execution but weakly incentivizing for leading low-carbon firms (Yunnan Aluminium, Hongqiao Yunnan) — low-carbon advantage gets partly offset by baseline decline. Versus EU EU ETS's "absolute-cap" approach — total cap declines yearly, all firms bid for total — directly rewards low-carbon. Chinese carbon market may gradually transition to a "baseline + absolute-cap" hybrid over 5-10 years, strengthening low-carbon incentive.

Chinese CEA price's medium-term trajectory is also key. Current CEA price only ~RMB 90/tonne — far below EU ETS's EUR 80-100 (RMB 600-750). CEA upward catalysts include: expanding more sectors, lowering free quota, bringing in financial institutions, building CCER-CEA inter-market linkage. By 2030 CEA likely climbs to RMB 150-180; by 2035 to RMB 200-250 — still well below EU ETS. The gap is the core barrier for Chinese aluminum CBAM compliance.

Beyond carbon, "green-power consumption" is also policy focus. NDRC/NEA's 2025 "Opinions on Deepening Green Electricity Trading Implementation" require electrolytic aluminum to reach 30 percent green-power consumption by 2030 and 50 percent by 2035. Pathways: direct green-power procurement via PPA or green-power exchange, green-power certificate purchase (decoupled from physical power), captive distributed PV/wind/storage builds. Current electrolytic-aluminum average green share ~22 percent — 8 ppt gap to 2030 target.

Green-aluminum certification: domestically dual-track — CNIA's "China Green Aluminium" voluntary (GREEN label) and internationally aligned ASI Performance Standard + Chain of Custody double certification. ASI is now in domestic use by Yunnan Aluminium, China Hongqiao, Nanshan Aluminum, Mingtai, Shenhuo as the key credential for serving Apple, Tesla, BMW, Daimler, Ford international auto and consumer electronics OEMs.

Recycled aluminum: 2024 MIIT Recycled Metals Industry Access Conditions (2024 Revision) reset environmental, energy, and scale thresholds for recycled-aluminum producers — single-line scale ≥50,000 tonnes/year, single-tonne energy ≤110 kg standard coal, comprehensive recovery rate ≥98 percent. The standard squeezes out small "scattered-disorganized-polluted" plants — concentration accelerates. The Finance Ministry maintains the 50-percent immediate-refund VAT preference (only for compliant firms) — a clear policy benefit for leading firms.

EU-China CBAM bilateral talks formally launched Q2 2026. First working-group meeting in Brussels focused on "carbon-cost reciprocity," "CEA-EU ETS quota two-way flow," and "Chinese aluminum exports CBAM filing simplification." This bilateral track is expected to run through mid-2027; meaningful breakthroughs face significant uncertainty. Even with partial recognition, given the gap, economic value is limited. Chinese aluminum should advance on three parallel tracks — domestic carbon-market price ascension + green-aluminum certification internationalization + overseas processing layout.

The "dual-control" energy framework deserves dedicated mention. Aluminum is one of NDRC-MIIT-MEE jointly defined "dual-high" (high-energy, high-emission) sectors, since 2021 included in "dual-control" management. Provinces impose annual targets — Shandong, Henan, Guangxi, Guizhou etc. set clear per-tonne aluminum energy-use ceilings (typically 13,500 kWh/tonne, with some H2 2025 lowering to 13,000). Dual-control directly drives cell-upscaling and energy-efficiency upgrade investment.

"Ultra-low emission" policy also accelerates. MEE's 2024 "Implementation Plan for Ultra-Low Emission Renovation in Key Sectors" requires electrolytic aluminum to complete ultra-low-emission renovation by 2027, with SO2, NOx, and PM concentrations at 30, 50, 5 mg/m3. Renovation CapEx ~RMB 5-8 million per 10,000-tonne capacity — sector cumulative renovation ~RMB 22-35 billion. Pressure is greater on coal-power-aluminum firms; hydropower aluminum mainly needs cell gas-collection upgrade and fluoride emission control.

The CBAM "embedded emissions" accounting also entails technical detail. CBAM filing requires direct emissions (production-process fossil-fuel combustion, process emissions), electric indirect emissions (external power carbon intensity at "origin actual" or "EU reference," whichever higher), and upstream-feedstock embedded emissions (alumina embedded, bauxite-shipping embedded). Electric indirect is the largest "deduction" for Chinese aluminum — China State Grid's average power carbon intensity is ~580 gCO2/kWh, far above the EU reference of 220 gCO2/kWh. This gap pushes Chinese electrolytic-aluminum CBAM filing values well above EU local equivalents. Response: apply for "actual plant power carbon intensity data" (Yunnan hydropower plants in particular), apply for indirect-emissions exemption, push EU on bilateral talks.

Resource and environmental taxes: bauxite resource tax has been ad-valorem since 2020 at 4-6 percent. Electrolytic-aluminum plant fluoride emissions, red-mud disposal, solid-waste disposal all attract environmental tax; coal-power-aluminum plants' SO2 and NOx are also subject. Combined resource + environmental tax adds RMB 180-220 per tonne aluminum — ~1 percent of sales.

"Green power consumption ratio" target also extends. Per NDRC/NEA's 2026 "Key Sector Green-Power Consumption Ratio Requirements," electrolytic aluminum is required to reach 25 percent green-power consumption by 2027, 40 percent by 2030, 55 percent by 2035. The requirement constrains coal-power-aluminum producers significantly — Shandong, Henan, Xinjiang etc. must achieve targets via green-power-certificate procurement or captive distributed wind-PV. The unified national green-power-certificate trading market launched Q2 2026 — single certificate = 1 MWh of green power, at RMB 20-50.

For mid-downstream chain, see aluminum curtain wall plants, aluminum window plants, aluminum formwork plants, and aluminum wheel plants — they reap policy upsides and also bear CBAM and carbon-market cost transmission.

12. Research institute's view — direction for the next three to five years

Building on the comprehensive analysis above, the Tianxia Gongchang Research Institute offers five core judgments for the next three to five years of Chinese aluminum.

First judgment: the 45-million-tonne ceiling will not be opened before 2028, but capacity relocation and indicator trading become a "grey market" normality. Under the hard ceiling, relocation to Yunnan, Sichuan, Inner Mongolia, and Xinjiang continues irreversibly. By 2028, Yunnan electrolytic-aluminum capacity may break 7.80 million tonnes, 17 percent national share. Xinjiang coal-power aluminum, under CBAM pressure, sees limited new capacity but remains a low-cost domestic source. Indicator prices likely hold the RMB 150-200 million per 10,000 tonnes range; should CBAM acceleration or national carbon-market expansion surprise, Shandong/Henan/Ningxia coal-power capacity may enter "accelerated exit" — indicator supply rises and center may shift down.

Second judgment: green aluminum (hydropower, wind-PV) premium over coal-power will widen further — from RMB 300-800 per tonne today to RMB 1,000-1,500. Key variable: EU ETS long-term ascension — 2030 EU ETS may hit EUR 200, 2035 over EUR 250. Coal-power EU exports then face implicit carbon tariffs of RMB 3,200-4,000; to offset, domestic carbon-market price will gradually ascend to RMB 150-180. Under this double pressure, green aluminum enjoys premium in export and earns "green priority procurement" policy benefits in domestic premium auto, consumer electronics, and battery sectors.

We must emphasize: the green-premium-widening judgment rests on three assumptions — strict CBAM execution, EU ETS price ascension, international auto procurement strictness. Any major reversal of these (e.g., EU pause on CBAM due to recession, ETS price decline from over-allocation, international auto procurement relaxation) could reverse the trend. The "premium widening" base case has ~60 percent probability; "flat" ~25 percent; "narrowing" ~15 percent.

Third judgment: recycled aluminum sees "three-stage growth" from 2026-2030 — Stage 1 (2026-2027) new scrap drives ~12-15 percent annual growth; Stage 2 (2028-2029) old scrap (especially NEV-scrap aluminum and end-of-life battery foil) drives 18-22 percent; Stage 3 (2030+) national carbon-market expansion and recycled substitution drives 22-26 percent. By 2030 recycled aluminum output exceeds 18 million tonnes, 28 percent share. CR10 may climb from 32 to 55-60 percent — Ye Chiu Metal, Sunbo Alloy, Xinjiang Joinworld, Mingtai will dominate.

Fourth judgment: battery foil enters "capacity race + tech upgrade" parallel phase. 2026 Chinese battery foil capacity reaches 1.10 million tonnes; actual output expected ~900,000 tonnes — utilization 82 percent. Dingsheng, Huafon, Shenhuo, Mingtai will hold >75 percent share. Tech upgrade vectors: ultra-thinning (12 to 8 micron), composite-coating, alloy optimization (Cu-Al composite current collector to mitigate thermal runaway). By 2028 Chinese battery-foil capacity over 1.50 million tonnes, output 1.30 million tonnes.

Fifth judgment: Chinese aluminum chain enters "two-way internationalization" — upstream to Guinea, Indonesia, India for alumina and electrolytic-aluminum; downstream to Europe, Southeast Asia, Mexico for processing capacity. Upstream: 2026-2030, Chinese aluminum firms add 4-5 million tonnes overseas electrolytic capacity (Indonesia, Guinea, Iran), add 7-9 million tonnes alumina (Indonesia, Guinea). Downstream: Chinese automotive sheet, battery foil, extrusion firms set up processing bases in Mexico, Hungary, Poland, Turkey, India, Malaysia — meeting local procurement of European, N. American, Southeast Asian OEMs and battery makers. This "upstream + downstream" two-way internationalization is the next-stage mainline of Chinese aluminum globalization.

The research institute further focuses on the "upstream-downstream synergy" opportunity. Yunnan Aluminium's Wenshan Funing plus the Qujing Malong, Wenshan Funing automotive-alloy die-casting park trials "smelter-melt-direct-supply-to-die-cast" integration — short-distance melt transport from smelter to die-cast plant, skipping ingot remelt, saving ~4.8 percent energy and ~6 percent carbon per tonne. This model will likely spread across Yunnan, Xinjiang, Inner Mongolia, and Sichuan smelter clusters over 3-5 years.

The institute also flags power-market reform impact on aluminum. The shift from "planned + market dual-track" to "fully market-based" trading puts electrolytic aluminum on "time-of-use tariff + green certification + ancillary services" composite pricing. Power costs may swing intraday or even hourly — demanding new procurement strategies and cell load-modulation capabilities. Leading firms (Yunnan Aluminium, China Hongqiao) have begun "dynamic cell-load modulation" systems — real-time cell power throttling (5-10 percent swing) tracking intraday price — to optimize power cost. This upgrade may become "new differentiator" in 5 years.

The institute also focuses on the "international competence" gap. Chinese aluminum has leading domestic-market scale, tech, and cost advantages, but international layout lags Rio Tinto, Alcoa, Hydro, and Hindalco. Chinese aluminum overseas capacity (bauxite, alumina, electrolytic, deep-processing combined) is only ~6 million tonnes — well below Rio Tinto's overseas total (e.g., 2.20 million tonnes electrolytic across five countries plus bauxite and alumina). The Chinese aluminum next-stage internationalization will likely focus on Indonesia, Guinea, Iran, Mexico, Turkey value-chain integration.

The institute also stresses long-term value of "green-aluminum brand" building. Yunnan Aluminium, China Hongqiao Yunnan Wenshan Funing, Shenhuo Yunnan — all at 2-4 tonnes per-tonne CO2 — are peer with Hydro and Rio Tinto BC Works in hydropower aluminum. But in international "green-aluminum brand recognition," Chinese Yunnan hydropower aluminum has not yet built a unified brand matrix akin to Norway's HyForge or Canada's RenewAl (Rio Tinto). Chinese aluminum's next internationalization stage should invest in "green-aluminum brand co-construction" — CNIA-led, with Yunnan Aluminium, China Hongqiao, Shenhuo, East Hope co-creating the "China Green Aluminium" international brand matrix to serve Apple, Tesla, BMW, Daimler, Dyson, Braun, Ford, Honda, Toyota global low-carbon supply chains.

The institute also issues the "old-scrap supply curve slope shift" judgment for ~2028. As large vehicle cohorts on roads from 2015-2017 enter 10-year scrap cycles, scrap-aluminum from autos surges 2025-2027. At 2015 China passenger-vehicle sales of 24 million units, single-unit aluminum 200 kg, full entry into 2025-2027 scrap could yield 4.80 million tonnes/year old-scrap; at 80 percent scrap rate and 90 percent recovery, ~3.50 million tonnes/year effective recycled feedstock — equivalent to 31 percent of 2025 Chinese recycled-aluminum supply. This curve shift is the strongest endogenous growth driver of recycled aluminum in coming three years.

The institute also flags "high-purity aluminum" subsegment's long-term value. 4N, 5N, 6N high-purity aluminum was historically for niche capacitor, electrode foil, sputter-target, mirror aluminum applications. But with advanced-node IC fab buildup, solid-state battery cathode current collectors, semiconductor sputter target localization — demand accelerates. By 2030 Chinese high-purity aluminum reaches 350,000 tonnes from current 120,000 (CAGR 19.4 percent). Xinjiang Joinworld (600888.SH), Dongyangguang Aluminium (600673.SH), Inner Mongolia Zhongyu High-Purity, and Changsha New Materials Research Institute's pilot lines are the major players. Xinjiang Joinworld, with Xinjiang Urumqi alloy electrolysis lines, produces ~28,000 tonnes 4N and ~6,000 tonnes 5N annually — the largest domestic player.

The institute also discusses "aluminum-plastic film" subsegment. Aluminum-plastic film is the outer material for soft-pack power and storage cells, composed of PE, foil, PA, PP multi-layers — foil being the core barrier. A soft-pack cell consumes ~0.04-0.06 m2 of aluminum-plastic film — ~3g aluminum. Chinese demand 2025 was ~700 million m2; 2030 expected at ~1.6 billion m2. The pain point is quality gap with Japan — Showa Denko, Dai Nippon Printing, Toppan Printing together hold ~75 percent of global high-end share. Local players Zijiang, Xinlun New Materials, Daoming Optics are accelerating localization, but high-end soft-pack power-cell aluminum-plastic film localization is only ~30 percent. This substitution process is expected to complete in 3-5 years.

Investment perspective: institute favors three lanes — Yunnan hydropower aluminum core names (Yunnan Aluminium, Shenhuo Yunnan); battery-foil leaders (Dingsheng, Huafon); recycled-aluminum integrators (Ye Chiu Metal, Sunbo Alloy). Cautious on: construction extrusion (property drag), coal-power aluminum (CBAM + carbon-market double pressure), independent alumina (Guinea dependency, high geopolitical risk).

Institute also notes "aluminum-copper substitution" marginal possibility. In NEV and battery, aluminum-copper substitution is a long-debated direction without near-term breakthrough. Aluminum current collector cuts cell per-Wh cost ~11-14 percent vs copper, but aluminum's conductivity is ~60 percent of copper, interface thermal stability is poorer, and reaction pathway with LFP cathode is unclear. BYD, CATL, EVE Energy are all testing "Cu-Al composite current collector" — Cu skeleton with Al surface — targeting conductivity-plus-lightweight balance. By ~2028, Cu-Al composite enters small-batch commercial use; 2030 market reaches ~300,000 tonnes scale.

The institute also notes profit-structure reshape for 2026-2030. Currently, profits concentrate upstream (smelter per-tonne RMB 1,800-2,800), with thin downstream (sheet GM 10-15 percent, extrusion 6-9 percent, foil 18-22 percent). As capacity-ceiling pressure rises and power-market reform drags upstream margins, downstream — pulled by NEV, battery, PV — see GM lift. This reshape makes mid-quality processors (Dingsheng, Mingtai, Huafon, Nanshan, Yongzhen) longer-term-valuable investment targets.

Looking at our 4.8-million-factory B2B platform's broader scope, our platform will keep tracking electrolytic, alumina, deep-processing, recycled, battery foil, and automotive sheet core lanes, providing precise structured factory retrieval for industry research, investment, sales prospecting, and supply-chain planning. See alumina equipment and electrolytic cell equipment factory maps for further details.

13. Risks — property decline, Guinea politics, CBAM ramp, hydropower variability

Any serious aluminum-industry report must seriously discuss risks. We list the top-five risks for 2026-2030 in order of probability and quantifiable loss.

Risk 1: construction aluminum demand weakness and property-chain decline. China's housing starts fell from the 2021 peak of 2 billion m2 to 900 million m2 in 2025 — a 55 percent five-year decline. Construction aluminum (windows, curtain walls, formwork) shrank correspondingly — 2021 building extrusion was 18 million tonnes; 2025 fell to 14 million, a 22.2 percent drop. Decline not yet bottomed — 2026-2027 housing starts expected at 700-800 million m2; building extrusion may drop another 10-15 percent. Pressure persists on small-and-mid extrusion firms in Shandong, Henan, Guangdong, Fujian, Jiangsu. Leading firms (Mingtai, Mindong Aluminium, Asia-Pacific) can partially offset via industrial-extrusion upgrade (PV frame, NEV structure, rail), but small-firm shakeout is inevitable.

Sector "shakeout" detail: 2023-2025 saw ~280 mid-small extrusion firms close or restructure, mostly in Foshan (Guangdong), Pizhou (Jiangsu), Zhengzhou (Henan), Zhangzhou (Fujian), Linyi (Shandong) clusters. 2026-2028 expects another 400-500 firm exits; CR20 may rise from current 41 percent to 58-62 percent. This "small-firm clearing + leader M&A" trajectory will define the next three years of Chinese aluminum deep-processing.

Risk 2: further Guinea export-policy tightening. 2024-2025 Guinea already twice raised export tariff and via Mining Code Amendment required domestic alumina-refining capacity. 2026-2028 potential variables: tariff to 20-25 percent; formal bauxite quota; mandatory in-Guinea alumina refining co-build by Chinese SMB, Boffa miners. Any one materially impacts global alumina — at Guinea's 70 percent share of Chinese imports, a 10 percent Guinea export reduction lifts Chinese alumina 12-18 percent. Hedges: increased Australian, Indonesian, Indian bauxite share; self-built Guinea alumina refineries (but 3-year build cycle limits near-term hedge).

Risk 3: CBAM ramp and scope expansion beyond expectations. Currently CBAM only covers primary aluminum and select intermediate processing. But EU signals possible expansion to all aluminum-content goods (autos, electronics, appliances). Then all Chinese aluminum-content exports to EU enter CBAM filing — heavy drag on Chinese auto, electronics, appliance exports. Electrolytic aluminum 2026-2030 CBAM ramp may accelerate — if EU ETS hits EUR 200 by 2030, per-tonne coal-power aluminum CBAM equivalent reaches EUR 3,200 — 60 percent of sales price — completely losing EU market.

Risk 4: hydropower aluminum's high sensitivity to drought. Yunnan and Sichuan — core hydropower aluminum carriers — depend on May-October rainy-season precipitation. 2023 Yunnan drought cut Jinsha hydropower output 31 percent; China Hongqiao's Yunnan Wenshan smelter cut 200,000 tonnes, Yunnan Aluminium triggered peak-shaving — single-firm loss ~RMB 1 billion. Under climate-change acceleration, extreme drought becomes more frequent. 2026-2030 if 2023-level drought recurs, Yunnan + Sichuan combined ~1.20-1.80 million tonnes forced cuts; supply-demand balance ruptures briefly; aluminum prices swing day-by-day.

Risk 5: national carbon-market expansion pace and price-ascension trajectory uncertainty. National carbon market expansion to electrolytic aluminum is a high-certainty policy landing, but specifics on timing, initial free quota, future tapering pace, and price ascension all carry uncertainty. If free-quota decline outpaces expectation, Shandong/Henan/Ningxia coal-power-aluminum players face surprise compliance cost pressure — some legacy capacity forced to exit early. Conversely, if CEA price ascension lags EU ETS, Chinese aluminum CBAM mapping difficulty rises, 2026-2030 Chinese aluminum exports to EU under sustained pressure.

Additional secondary risks: global geopolitical tension may disrupt aluminum shipping channels (West Africa to China, Guinea to China); global data-center compute demand marginal slowdown could pull near-term aluminum down; if Chinese NEV sales growth slows below 10 percent annual, automotive sheet and battery foil demand weakens; ASI standard iteration may strand some Chinese firms in certification; Russian aluminum sanctions push large discounted volume to Chinese market, suppressing domestic prices. Individually low probability, but combined concurrent realization adds sustained pressure.

Risk transmission: five risks are not isolated — they form "risk resonance." For instance, Guinea export tightening (Risk 2) combined with Yunnan drought (Risk 4) creates simultaneous upstream alumina and downstream electrolytic constraint — short-term price spike probability rises. CBAM ramp (Risk 3) plus national carbon-market expansion (Risk 5) generates "carbon-cost dual squeeze" on coal-power, forcing Shandong/Henan/Ningxia legacy firms to exit within 2-3 years. These resonance scenarios aren't certain to occur but have structural implications.

Plant resilience: leading firms (Chalco, China Hongqiao, Yunnan Aluminium) with scale, tech, brand, green-power, overseas-resource advantages cope well with single risks. Specialized mid-size firms (Dingsheng, Mingtai, Huafon) rely on single-segment health — if primary product (battery foil, automotive sheet) weakens, performance dives quickly. Small extrusion firms are mostly "low-margin, high-operating-leverage" — any adverse change (aluminum price, demand, energy use, environmental) drives losses or bankruptcy.

Hedging tools: Chinese firms widely use SHFE futures, options, calendar-spread arbitrage for price risk; alumina and bauxite long-term contracts for raw-material risk; PPA and self-built power for power risk; export hedging combinations for FX. But for geopolitics (Guinea export, CBAM ramp, Russian sanctions outcomes) and climate-driven hydropower variability, financial hedging is poor — primary reliance is physical supply backup and diversified deployment.

Risk 6 (secondary but not negligible): emerging-country aluminum producers' anti-dumping cases on Chinese aluminum. India, Turkey, Mexico, Brazil, Vietnam, Indonesia have all launched recent anti-dumping investigations against Chinese aluminum foil, extrusion, and sheet. India MoC launched a new round of aluminum foil anti-dumping in September 2025 with initial rates 19-26 percent; Turkey Ministry of Economy made a preliminary decision in March 2026 with 21 percent anti-dumping; Mexico Ministry of Economy made initial decision in May 2026 with 17-23 percent on Chinese sheet. Combined, these affect ~1.20 million tonnes/year of Chinese aluminum exports.

Risk 7: Indonesia, Vietnam, India, etc. emerging aluminum-producers' acceleration of domestic value chains substitutes Chinese exports. Indonesia plans 2030 electrolytic 2.50 million + alumina 20 million; Vietnam plans 1.20 million + 10 million; India plans 6 million + 15 million. Local capacity expansion gradually substitutes their current Chinese-aluminum imports. Chinese aluminum exports to Indonesia, Vietnam, India see clear downside in 2028-2030. Chinese aluminum response: "early overseas layout" — joint-ventures or wholly-owned plants converting "export trade" to "investment trade."

Risk 6 secondary refinement: India aluminum capacity expansion brings "value-chain coordination" imbalance. Hindalco + Vedanta combined electrolytic from 3.78 million tonnes 2025 to 6 million by 2030; Indian alumina from 7.90 million to 15 million. Indian bauxite import demand thus rises sharply, from <10 million tonnes today to 30 million by 2030. India sources bauxite from Guinea, Indonesia, Australia — heavily overlapping with China. Sino-Indian Guinea bauxite "contest" is among the most-watched global geopolitical dynamics over 5-10 years.

Risk 9: battery-foil overcapacity and "involution" pressure. 2023-2026 saw aggressive battery-foil capacity expansion in expectation of strong NEV demand — Dingsheng, Huafon, Shenhuo, Mingtai, Wanshun, Chalco Luoyang Copper added ~800,000 tonnes cumulatively; 2026 Chinese battery foil effective capacity reaches 1.10 million tonnes. But H2 2025 saw NEV growth slow (2023's 36 percent to 2025's 19 percent), with battery-cell inventory pressure. Battery-foil prices fell from 2023's RMB 45,000/tonne to Q2 2026's RMB 34,000 (-24.4 percent). Battery-foil GM fell from 2023's 25 percent to Q2 2026's 16 percent. If capacity outruns demand growth, 2027-2028 may enter "capacity excess + price war" shakeout.

Risk 10: tech-substitution risk. Long-term, aluminum faces two-tier tech-substitution risk: Tier 1 — carbon fiber, magnesium alloy, titanium alloy partially substituting aluminum in aerospace, premium auto, electronics; Tier 2 — solid-state batteries, aluminum-air batteries disrupting battery-foil demand. Tier 1 substitution is slow — carbon fiber, magnesium, titanium combined costs remain well above aluminum, so substitution proceeds gradually. Tier 2 is more uncertain — solid-state batteries' commercialization may change battery-foil demand structure (still uses foil but with different spec) — requires continuous monitoring.

Risk 11: synthetic-biology / novel-aluminum-alloy "disruptive-tech risk." Frontier science is exploring synthetic biology (bacterial bioleaching, yeast metabolic engineering) to recover high-purity aluminum from scrap, plus novel alloys (high-entropy aluminum, aluminum-based metallic glass, aluminum-ceramic composites). If these break through, 10-15 year window could disrupt traditional aluminum chain. But currently at lab stage — at least 10+ years from industrial application. Chinese aluminum should track and modestly invest, without overreacting to near-term impact.

For the plant landscape under these risks, see aluminum furniture plants, aluminum heat-sink, and aluminum-clad steel — including diversified leaders with strong cycle-resistance and small firms sensitive to specific risks.

14. Data sources, references, and platform retrieval notes

The report's core data and judgments come from:

Chinese national and industry-level data: NDRC's "2024 Electrolytic Aluminum Industry Capacity and Operation Report," "China Electrolytic Aluminum Industry Policy Window Continuation Opinions"; MIIT's "Aluminum Industry High-Quality Development Plan (2024-2027)," "Recycled Metals Industry Access Conditions (2024 Revision)"; CNIA's "2025 China Aluminum Industry Statistical Yearbook," "China Electrolytic Aluminum Industry Green and Low-Carbon Development Report (2025)"; General Administration of Customs of China "2025 China Aluminum and Products Trade Analysis"; MEE "2024 National Carbon Market Annual Report."

Listed-company annuals: Chalco 601600.SH 2025 annual, Yunnan Aluminium 000807.SZ 2025 annual, Tianshan Aluminium 002532.SZ 2025 annual, Shenhuo 000933.SZ 2025 annual, Nanshan Aluminum 600219.SH 2025 annual, Mingtai 601677.SH 2025 annual, China Hongqiao 1378.HK 2025 annual, Dingsheng 603876.SH 2025 annual, Ye Chiu Metal 601388.SH 2025 annual, Sunbo Alloy 002996.SZ 2025 annual.

International data and institutional reports: Rio Tinto 2025 integrated annual + aluminum-segment, Alcoa Corporation 2025 10-K, Norsk Hydro 2025 integrated annual, Hindalco / Aditya Birla 2025 integrated annual, Emirates Global Aluminium 2025 sustainability report, Rusal 2025 integrated annual, Vedanta Aluminium 2025 integrated annual, IAI "World Aluminium Statistics 2025" and "Global Aluminium Cycle 2025," CRU Aluminium Outlook 2026 H1, CRU Aluminium Quarterly 2026 Q2, Wood Mackenzie Aluminium Service 2026 Q1 Strategic Issue, BloombergNEF "Sustainable Aluminium Pathway 2026," "Global Aluminium Outlook 2026 H2," Reuters, Nikkei thematic reports.

Policy and regulatory texts: EU CBAM Regulation EU 2023/956 and implementing rules, EU ETS 2026 annual quota notice, European Commission "CBAM Implementation Rules Amendment Draft" 2026 edition, Guinea Mining Code Amendment 2025, Indonesia Mining Resource Export Management Regulation No. 2 of 2023, China MEE "National Carbon Emission Trading Market Sector Expansion Implementation Plan (2025-2030)," Canada Climate Protection Act Amendment 2026, US Clean Competition Act Draft.

Chinese Ministry of Commerce, GAC monthly statistics: This report draws on monthly trade data, EU export structure, and Indian anti-dumping data from these departments. Chinese industry-association research subsidiaries: CNIA Aluminum Branch "2025 China Aluminum Industry Monthly Data Compilation," CAAM "2025 China NEV Production-Sales Monthly Data," CPIA "2025 China PV Module Shipment Monthly Data," CABIA "2025 China Power-Battery Installation Monthly Data," etc.

International institutional data supplement: World Bank "Global Commodity Markets Outlook 2026," IMF "World Economic Outlook 2026," OECD "Trade and Commodities Review 2026," IEA "World Energy Outlook 2026," UNCTAD "Commodity Trade Statistics 2025," WTO "Aluminium Trade Review 2026."

Industry research institutions and media: This report also references CNIA "China Aluminium Industry Yearbook 2025," CNIA Recycled Metals Division "2025 China Recycled Aluminium Industry Development Report," SMM "2026 China Aluminium Value Chain Spring Report," Antaike "2026 China Electrolytic Aluminium Industry Research Report," Mysteel "2026 Aluminium Value Chain Survey Report," CITIC Securities "China Aluminium 2026 Annual Strategy," CICC "2026 Global Aluminium Outlook," Huatai Securities "Nonferrous 2026 Mid-Year Strategy — Aluminium," Haitong Securities "2026 China Electrolytic Aluminium Industry Deep Report," GF Securities "NEV Aluminium Consumption Deep Dive — Battery Foil, Automotive Sheet, Lightweight Structures."

Academic references: process detail, cell parameters, inert-anode progress, recycled-aluminum process — drawn from "Metallurgical Industry Press · Aluminium Metallurgy" (textbook), Journal of Aluminium Smelter Technology, Light Metals (TMS annual proceedings), Chinese Journal of Nonferrous Metals, Aluminium International Today, Light Metal Age, Aluminium Practice and other specialized journals and proceedings.

Data baseline: report core data baseline June 20, 2026; financial data are 2025 full-year consolidated; Q1 2026 data drawn from listed-company Q1 reports and industry monthly data. All prices reflect average transaction prices, in original currencies; conversions use the People's Bank of China central parity as of June 20, 2026, where needed. FX: PBoC central parity June 20 2026 USD/CNY 7.0584, EUR/CNY 7.6512, JPY/CNY 0.0436, GBP/CNY 8.8932, CAD/CNY 5.0814, AUD/CNY 4.5891, INR/CNY 0.0823.

Plant-landscape data source: this report's plant-distribution data come from the Tianxia Gongchang 4.8-million-factory structured database, integrating business registration, tax, value-chain, customs, industry park, environmental, and electricity multi-source data, with AI-driven process/product classification and on-site verification sampling — covering 1,965 sub-categories. Readers may search at our platform (tianxiagongchang.com) by process, region, capacity, downstream application.

Special note: "real producing factories" in this report refers to legal entities still operating with identifiable process/product categories — excluding pure traders, agents, decoration outfits, or equipment dealers. This is our platform's core differentiator versus traditional business-registration services — the latter shows registration data only, without identifying actual operating status. Our platform's identification capability rests on years of cross-validation across business, tax, customs, public tender, environmental, and electricity multi-source data, plus AI classification and on-site sampling, with precision improving year by year.

Data-refresh mechanism: the report's plant data are baseline June 20, 2026, but the platform's database is continuously updated — ~5,000-8,000 new factory registrations weekly, ~20,000 product/capacity/operating-state updates monthly, quarterly nationwide operating-state verification sampling. Readers accessing the platform at different times see dynamic changes in the factory map. The specific plant counts in this report (e.g., ~30 alumina refineries, ~150 sheet-and-plate plants) are baseline-instantaneous statistics; actual counts will fluctuate.

Data limitations: company-financial figures (per-tonne profit, combined cost, GMs) integrate listed-company annuals, association surveys, third-party estimates. Different sources may use slightly different bases — some details may not exactly align with single-source disclosures. Readers citing should cross-verify against the original source for the use case. Time framings such as "H2 2026," "next 3-5 years," "around 2030" reflect baseline judgment of June 20, 2026 — actual evolution may diverge.

Report writing also drew on interviews with senior practitioners across the chain — bauxite mining engineers, alumina refinery process leads, electrolytic-aluminum cell engineers, deep-processing rolling engineers, recycled-aluminum alloy-blending engineers, aluminum die-casting mould developers, automotive OEM alloy procurement leads, NEV battery-foil procurement engineers. These ground-truth perspectives are crucial. We gratefully acknowledge their open contribution.

Report also draws on global-consulting research on Chinese aluminum — McKinsey "2026 Global Lightweighting Materials Outlook," BCG "2026 China NEV Aluminum Consumption Deep Report," Bain "2026 Global Green Aluminium Value Chain Layout," Roland Berger "2026 China Recycled Metals Industry Research" — providing complementary strategic and operational perspectives.

Complete value-chain factory-retrieval keyword list: see aluminum alloy profile, aluminum die-cast parts, automotive aluminum alloy, aluminum alloy ingot, battery aluminum foil, aluminum composite panel for real factory distributions.

Global recycled-aluminum "long-tail sources" also deserve brief mention. Beyond China, EU, US, Japan, the four recycled-aluminum heartlands, Korea, Mexico, Brazil, Turkey, India, and Southeast Asia are noteworthy. Korea produces ~1.20 million tonnes annual recycled aluminum, serving Korean automotive industry; Mexico ~900,000 tonnes serving N. American autos; Brazil ~1.40 million serving S. American autos and construction; Turkey ~800,000 tonnes serving European and Middle Eastern autos. These regions' recycled-aluminum capacity expansion is steady but consistent — projected global recycled-aluminum 2030 output of 48 million tonnes versus today's 32 million.

Copyright is owned by our research institute. Unauthorized commercial use prohibited. Industry institutions citing core data and judgments please credit "Our research institute, 'China Aluminum Smelting 2026 — Alumina Value-Chain Integration and Electrolytic-Aluminum Green-Power Revolution.'"