# From ground to market: Turning rock into metal: concentration, smelting and refining Source page: https://www.materialinsights.org/learn/commodities/from-ground-to-market/p3/ [Understanding Commodities](https://www.materialinsights.org/learn/commodities/) / [Module 2: From ground to market](https://www.materialinsights.org/learn/commodities/from-ground-to-market/) Part 3 of 5 ## Turning rock into metal: concentration, smelting and refining Almost nothing leaves a mine in saleable form. Ore is crushed and concentrated near the mine, then smelted and refined, often in another country. Each step sheds mass, uses energy and supports a business with its own margins. This part follows copper, iron and aluminium through those steps and ends with the arithmetic of how a custom smelter is paid. ### 3.1 Beneficiation concentrates ore so the costly steps handle less rock Beneficiation is the physical upgrading done before any chemistry. Tilton and Guzmán describe the sequence for copper sulphide ore. The ore is crushed and then ground to a fine powder, so that each mineral grain is freed from its neighbours. In froth flotation the slurry is mixed with reagents that make the copper minerals repel water; air bubbles carry them to the surface as a froth, while the barren gangue sinks. The product is a concentrate of 25 to 35 per cent copper. The logic is energy. Skinner's point, as Tilton reports it, is that crushing and flotation are cheap mechanical steps, so the energy-hungry work of breaking chemical bonds is applied only to the concentrate. Radetzki and Wårell note that copper loses about two-thirds of its weight between concentrate and refined metal. That favours refining near the mine, but power costs, scale and trade policy often pull the smelter elsewhere. Iron ore is beneficiated differently. Smil explains that blast furnaces need a rich burden, preferably above 58 per cent iron, with few fines below 5 millimetres, because fines block the gas rising through the furnace. Chinese ores average only 30 to 40 per cent iron against 62 per cent in Australia, so they need far more work. Low-grade ore is milled, washed and magnetically separated, sometimes floated to strip silica, then agglomerated. Sinter, fired at the steelworks from fines, coke breeze and flux, feeds about 70 per cent of the world's hot metal but crumbles in transit. Pellets, fired balls of concentrate holding 60 to 65 per cent iron, are hard enough to ship. - Tilton & Guzmán (2016), Mineral Economics and Policy · pdf p. 162 - Tilton (2003), On Borrowed Time? · pdf p. 79 - Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf p. 79 - Smil (2016), Still the Iron Age · pdf pp. 124-128 ### 3.2 Pyrometallurgy and hydrometallurgy are two roads to pure copper Copper concentrate takes the hot road. In the smelter the sulphide is melted and oxidised; iron goes into slag and sulphur leaves as sulphur dioxide. The product is blister or anode copper of 97 to 99 per cent. Electrolytic refining then dissolves the anodes and plates cathode of 99.99 per cent (Tilton and Guzmán). This is pyrometallurgy, and its great environmental problem is the sulphur dioxide and arsenic in the off-gas. Tilton reports that the Chuquicamata smelter raised sulphur dioxide capture from zero to 80 per cent, and arsenic capture from 35 to 90 per cent, between 1980 and 1999, at a cost of more than $600 million. The best Japanese smelters capture all but 1 or 2 per cent. Oxide ores take the wet road. They are leached with weak acid. In solvent extraction an organic chemical pulls copper out of the dilute leach liquor and is then stripped into a clean, concentrated electrolyte. Electrowinning plates cathode directly. SX-EW skips the concentrator and the smelter, and often the mine itself, since old dumps can be leached. Tilton put it at about a quarter of world copper in the early 2000s, with no sulphur dioxide emitted at all. Hydrometallurgy reaches well beyond copper. High-pressure acid leaching treats nickel laterites. Rare earths are separated from one another by long chains of solvent-extraction steps, which Pitron describes as repetitive and acid-heavy, using at least 200 cubic metres of water per tonne, water rarely treated in China. The chemistry that frees a metal from rock also creates wastes that someone must manage or ignore. Case #### How innovation rescued US copper Tilton and Guzmán describe a near-death. Between 1970 and 1985 US copper output fell by nearly a third, its share of Western output fell from 30 to 17 per cent, and employment dropped 70 per cent. Business Week declared US mining dead. Yet by 1995 output was 72 per cent above its 1985 level, and labour productivity in 2001 was triple that of 1980. Grades did not help: they fell from 0.78 to 0.60 per cent over 1971 to 1993. Three-quarters of the productivity gain came within existing mines, from SX-EW, bigger trucks, in-pit crushers, flexible work rules and better pit design. - Tilton & Guzmán (2016), Mineral Economics and Policy · pdf pp. 161-163 - Tilton (2003), On Borrowed Time? · pdf pp. 101, 103 - Humphreys (2015), The Remaking of the Mining Industry · pdf p. 232 - Pitron (2020), The Rare Metals War · pdf pp. 26-27 ### 3.3 Iron and aluminium show how energy shapes the metallurgical route Steel has two main routes. In the integrated route a blast furnace reduces iron ore with carbon. Smil, citing World Steel Association figures, gives typical inputs for a tonne of steel made this way of about 1,400 kilograms of iron ore, 800 of coal (most of it turned into coke) and 300 of limestone and other flux. Coke both heats the furnace and strips oxygen from the ore, so carbon dioxide comes from the chemistry, not just the fuel. A basic oxygen furnace then blows oxygen through the molten iron to burn off excess carbon. The best blast furnaces need about 12 GJ per tonne of hot metal, and integrated steel needs around 20 GJ a tonne in all. The electric arc furnace melts scrap instead, using about 2.5 GJ of electricity per tonne, roughly 6 GJ in primary energy (Smil). Around 2015 EAFs made nearly 30 per cent of the world's steel, about 60 per cent of US steel and over 90 per cent in the Middle East, where they also melt direct-reduced iron made with cheap natural gas. Aluminium is the extreme case. Bauxite is refined into alumina, and USGS gives a rule of thumb of 4 tonnes of bauxite to 2 of alumina to 1 of metal. The Hall-Héroult process dissolves alumina in a molten salt bath and splits it with direct current, consuming carbon anodes. Smil puts aluminium at about 175 GJ per tonne, mostly electricity, against 45 for copper and under 20 for steel. Humphreys (2015) puts carbon dioxide at about 20 tonnes per tonne of aluminium on coal power and nearer 10 on hydro, against about 2 for steel. Tilton (2003) cites a lower coal-power estimate of nearly 17 tonnes per tonne. Humphreys adds that energy makes up around a quarter of the cost of producing metals. Case #### Japan's vanishing aluminium industry Tilton and Guzmán describe how Japan built one of the world's largest aluminium smelting industries in the 1960s, powered largely by imported oil. When oil prices jumped in the 1970s, smelting in Japan stopped making sense, and the country dismantled most of the industry within a decade. China later did the opposite. Humphreys notes that China lacked good bauxite and cheap power, yet low plant costs and subsidised or stranded hydropower made it 46 per cent of world output by 2013. Open the Aluminium tile: USGS puts China at 44 of 72.8 million tonnes in 2024, about 60 per cent, and the tile shows an HHI above 3,700. - Smil (2016), Still the Iron Age · pdf pp. 93, 102, 109, 131, 145, 150-155 - USGS Mineral Commodity Summaries 2026, Bauxite and Alumina; Aluminum [web ↗](https://pubs.usgs.gov/publication/mcs2026) · footnotes; World Smelter Production - Humphreys (2015), The Remaking of the Mining Industry · pdf pp. 81-82, 111, 234 - Tilton & Guzmán (2016), Mineral Economics and Policy · pdf p. 69 - Tilton (2003), On Borrowed Time? · pdf p. 101 ### 3.4 A custom smelter earns a fee that rises and falls with concentrate supply Much copper concentrate is sold to custom smelters that own no mines. Humphreys explains how they are paid. The buyer negotiates a discount on the value of the copper contained in the concentrate, the treatment and refining charge or TC/RC, to cover the cost of turning it into cathode. The smelter buys concentrate at the metal price less that discount, sells cathode at the full price, and lives on the difference after its own costs. A worked example shows the stakes. Take one dry tonne of 30 per cent concentrate, containing 300 kilograms of copper. At the September 2026 price of $14,474 a tonne, that copper is worth about $4,340. Over 1980 to 2005 the discount averaged around 23 per cent of the copper price, about $1,000 on this tonne. In 2007 it fell to about 5 per cent, roughly $220. Smelting costs did not fall by three-quarters, so the business became unprofitable for most smelters, and the charge stayed low until new mine supply eased the concentrate market in 2013. Why did it fall? The fee is a price for smelting capacity, set by the balance between concentrate and furnaces. China built smelters faster than the world built mines. Radetzki and Wårell add that VAT rebates let Chinese smelters offer low charges for toll smelting, which expanded Chinese capacity while smelting stagnated elsewhere. Pirrong notes a twist: the copper in concentrate can be hedged on the LME, but the treatment charge cannot. When smelters compete for scarce feed, value shifts to miners. When mines outrun smelters, it shifts back. In the data #### Mine here, smelt there USGS figures for 2024 show how far mining and refining have separated. Chile mined 5.51 million tonnes of copper but refined 1.94 million. China mined 1.84 million tonnes but refined 12.4 million, about 45 per cent of the world's 27.6 million, and the 2025 estimate rises to 14 of 29 million. Open the Copper tile and compare lenses: mine output is spread out, with an HHI near 1,090, while China took 63 per cent of world imports of copper ore and concentrate by value in 2024. Chinese demand for smelter feed largely sets the TC/RC. - Humphreys (2015), The Remaking of the Mining Industry · pdf pp. 79-80 - Radetzki & Wårell (2021), A Handbook of Primary Commodities · pdf p. 81 - Pirrong (2014), The Economics of Commodity Trading Firms · pdf p. 23 - USGS Mineral Commodity Summaries 2026, Copper [web ↗](https://pubs.usgs.gov/publication/mcs2026) · World Mine and Refinery Production ### See it in the table [H₂O Water](https://www.materialinsights.org/commodities/water/)[Gas Natural gas](https://www.materialinsights.org/energy/natural_gas/)[Cu Copper](https://www.materialinsights.org/commodities/copper/)[Al Aluminium](https://www.materialinsights.org/commodities/aluminum/)[Ni Nickel](https://www.materialinsights.org/commodities/nickel/)[Fe Iron ore](https://www.materialinsights.org/commodities/iron_ore/)[St Steel](https://www.materialinsights.org/commodities/steel/)[REE Rare earths](https://www.materialinsights.org/commodities/rare_earths/)[As Arsenic](https://www.materialinsights.org/commodities/arsenic/)[NaCl Salt](https://www.materialinsights.org/commodities/salt/)[O2 Oxygen](https://www.materialinsights.org/chemicals/oxygen/)[CO2 Carbon dioxide](https://www.materialinsights.org/chemicals/carbon_dioxide/) ### Check your understanding Why are copper sulphide ores concentrated by flotation before smelting? Skinner's argument, reported by Tilton, is that mechanical concentration lets the costly chemical steps work on a small fraction of the mined tonnage. In 2007 the copper TC/RC fell to about 5 per cent of the copper price, from a 1980-2005 average of about 23 per cent. Who gained? A smaller discount means the miner receives more of the contained metal's value. Humphreys says the change made smelting unprofitable for most of the industry. Which statement about the electric arc furnace is correct? Smil gives about 2.5 GJ per tonne of direct electricity for an EAF against roughly 20 GJ for integrated steel. The EAF dates from Siemens and Héroult in the late nineteenth century. [← Part 2: From discovery to closure: the mine life cycle and how ore is extracted](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p2/)[Mark read · Part 4: Hydrocarbons and harvests: the other extraction chains →](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p4/) ## Other sections - [Module overview](https://www.materialinsights.org/learn/commodities/from-ground-to-market/index.md) - [What the ground offers: deposits, grades and waste](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p1/index.md) - [From discovery to closure: the mine life cycle and how ore is extracted](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p2/index.md) - [Hydrocarbons and harvests: the other extraction chains](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p4/index.md) - [Where value sits, and the licence to keep digging](https://www.materialinsights.org/learn/commodities/from-ground-to-market/p5/index.md) - [Key terms, quiz and sources](https://www.materialinsights.org/learn/commodities/from-ground-to-market/review/index.md)