Skip to content

XLSLIDES COPPER RESEARCH ATLAS · 2025 REFERENCE MODEL

Global copper map.
From the pit to the wire.

The mine. The smelter.
The refinery. The road to the coast.

Explore where copper is mined, smelted, refined and shipped, and the estimates connecting them. Select a record to inspect its source, period, stage and verification status.

Explore the copper map Free to explore · No sign-up · Built 13 November 2026
250MINE RECORDS
179SMELTER & REFINERY RECORDS
98PORT, RAIL & WAREHOUSE RECORDS
267FLOW RECORDS
READ THE PERIOD, STAGE & BASIS

94 of 250 mine records carry a single-site 2025 copper figure. Group totals, ore tonnages and “included elsewhere” texts are kept as written but not counted, and 609 site records carry a research flag on at least one field. Inspect the coverage notes

THE INTERACTIVE RESEARCH ATLAS

Follow the tonne. Inspect the estimate.

Open full explorer
TRY A STARTING POINTLOADING ATLAS
Overview drawn from the atlas geography
Loading the atlas…One research snapshot. No sign-up.
Drag to pan · use + / − to zoom · search or select a record2025 reference model · Lines join representative points, not traced shipments

ONE METAL. MANY STEPS.

Keep the chain
in view.

  1. 01

    At the mine

    250 mine records and 76 projects. Read output as contained copper, and check the stage: copper in concentrate, SX-EW cathode or a group total.

  2. 02

    At the smelter and refinery

    116 smelter and acid-plant records and 63 refinery and SX-EW records. Output and nameplate capacity are different numbers, and many plants run on imported concentrate.

  3. 03

    Along the route

    267 flow records, 169 of them drawn, and 98 ports, rail lines and warehouses. A line is a documented relationship or trade total, not a traced cargo.

WHERE COPPER IS MINED

Mining is concentrated.
Named records cover part of it.

Country mine output for 2025 as estimated by the USGS and transcribed by the research (world about 23,000 kt). The darker bar is the sum of the model’s single-site mine figures in that country: a coverage indicator that mixes periods and stages, not a national total.

Chile
5,300 kt
DR Congo
3,200 kt
Peru
2,700 kt
China
1,800 kt
United States
1,000 kt
Russia
1,300 kt
Indonesia
710 kt
Zambia
940 kt
Australia
730 kt
Kazakhstan
710 kt
Mexico
690 kt
Canada
500 kt
Poland
410 kt
India
23 kt

USGS world 2025e: 23,000 kt, including about 3,000 kt from other countries · ICSG: about 23.2 Mt · model single-site sum: 14,324 kt from 94 records

How the model covers each country

  • Chile 4,614 kt from 23 single-site mine records against 5,300 kt (USGS); 41 mine and project records in all.
  • Democratic Republic of the Congo 1,489 kt from 4 single-site mine records against 3,200 kt (USGS); 20 mine and project records in all.
  • Peru 2,523 kt from 10 single-site mine records against 2,700 kt (USGS); 25 mine and project records in all.
  • China 350 kt from 2 single-site mine records against 1,800 kt (USGS); 15 mine and project records in all.
  • United States 849 kt from 8 single-site mine records against 1,000 kt (USGS); 27 mine and project records in all.
  • Russia 70 kt from 1 single-site mine record against 1,300 kt (USGS); 15 mine and project records in all.
  • Indonesia 555 kt from 2 single-site mine records against 710 kt (USGS); 9 mine and project records in all.
  • Zambia 601 kt from 4 single-site mine records against 940 kt (USGS); 10 mine and project records in all.

Largest single-site mine figures

MinektAs recorded
EscondidaChile1,3451345.1
Tenke Fungurume (TFM)Democratic Republic of the Congo741741.1 (TFM+KFM combined; split not found)
Grasberg Block Cave (GBC)Indonesia460~460 (PTFI district total, 1,015 Mlb) (modelled)
Las Bambas (Ferrobamba / Chalcobamba pits)Peru411410.8
Buenavista del Cobre (Cananea) open pitMexico409409.4 (incl. SX-EW; Northern Miner ranking)
Cerro VerdePeru408408 (modelled)
CollahuasiChile406406
Kakula mine (Kamoa-Kakula)Democratic Republic of the Congo389388.8 (whole complex, Cu in concentrate)
AntaminaPeru382381.8
Oyu Tolgoi - Oyut open pit and concentratorMongolia345345 (OT total, 100%)
Morenci mineUnited States318317.5
QuellavecoPeru310310.2

Largest smelters and refineries

PlantktBasis
Jinchuan Group copper smelter and refinery (Jinchang)Refinery · China800capacity: smelter 400; refinery 800 (ICSG 2025)
LS MnM Onsan refineryRefinery · South Korea680capacity: 680
Guangxi Nanguo Copper (Nanfang Nonferrous)Smelter · China675capacity: 675 (ICSG 2025); 300 phase 1 (WoodMac)
Daye Nonferrous smelter/refinery (Huangshi)Refinery · China600capacity: refinery 600
Dongying Fangyuan NonferrousSmelter · China550output: ~550 (flagged for checking)
LS MnM Onsan smelterSmelter · South Korea550output: ~550-600 (cathode; see note)
Chifeng Jintong CopperSmelter · China520capacity: 520 (ICSG 2025)
Guixi Smelter (smelter + refinery)Smelter · China520capacity: smelter 520 (ICSG 2025); refinery 1,100
Yunnan Copper (Southwest Copper) KunmingRefinery · China500capacity: 500 (ICSG 2020)
Kamoa-Kakula copper smelterSmelter · Democratic Republic of the Congo500capacity: 500

COVERAGE & EVIDENCE

A record is a starting point.

The atlas keeps the owner’s wording, mixed periods and the research’s own verification flags. A number is read only where the text gives the record’s own copper. Detailed coordinates do not establish complete coverage or a traced supply chain.

How output values were read

  • No output recorded447
  • Own single-site figure131
  • Wording without a figure22
  • Group, company or country total19
  • Counted in another record15
  • Not a copper tonnage (ore, cargo, other metal)12
  • Covers linked operations7
  • Residual for several producers1

Counts of mine, plant, fabrication and company records, not shares of tonnes. Ports, rail lines, warehouses and desalination plants keep their wording only.

Site confidence

  • high98
  • medium265
  • low414

Records by layer

LayerRecords
Mines250
Projects & artisanal areas76
Concentrators & tailings33
Smelters & acid plants116
Refineries & SX-EW plants63
Rod mills & fabricators73
Ports, rail & warehouses98
Desalination plants10
Company headquarters58

267 flows (169 drawn), 241 money items and 120 facts complete the snapshot. Money items and facts have no coordinates and appear as tables in the explorer.

Research flags and modelled values

  • Site records with a flagged field609
  • Flows with a flagged field100
  • Money items with a flagged field56
  • Records with modelled values251
  • Sites citing unverified background knowledge42
Download snapshot · JSON · 2.2 MB

Limits to retain

  • This is a selected documentary model of 777 site records, 267 flows, 241 money items and 120 facts, not a complete or live inventory of the copper industry.
  • Only 94 of 250 mine records carry a single-site 2025 copper figure. Their sum, 14,323.6 kt, is a coverage indicator for named records, not a world total: ICSG puts 2025 world mine output at about 23.2 Mt and the USGS at about 23,000 kt.
  • Output figures mix calendar 2025, fiscal years ending in 2025 or 2026, guidance, part-year and earlier-year values, and different stages: copper in concentrate, SX-EW cathode, anode or refined copper. The selected panel keeps the owner’s wording, period and stage.
  • Group, company, district and country totals are kept as written but are not used for symbol size or sums; records described as “included in” another record carry no number of their own. Ore, cargo, gross-concentrate and copper-equivalent tonnages are shown as text only.
  • 45 site records, 100 flows and 101 money items contain a value or note the research marked as modelled. Modelled values are the research’s own estimates, often a share of a group total or a conversion at the 2025 LME average of US$9,939 a tonne.
  • The research flags 609 site records, 100 flows and 56 money items for checking in at least one field, most often notes, owner stakes and start years. 42 site records cite background knowledge that was not verified. The selected panel names the flagged fields.
  • Flow lines connect representative points: a named site where the endpoint matches one, otherwise a country hub. 98 flows with an unstated, worldwide or same-place endpoint are listed in the table but not drawn. A line is a documented relationship or trade total, not a traced shipment.
  • Money items are payments, revenues, deals and costs with recipients but no coordinates; they appear as tables. Items differ in year, currency conversion and basis (gross, net, attributable), so they should not be added together.
  • Coordinates are often approximate (pit, plant or city level), and some records describe closed, historic or planned assets. Capacities are design or nameplate figures and differ between sources; symbol size is only a display scale.

Start with a tonne and a question

A copper map is most useful when a location leads to a question you can check. A mine can sit beside a smelter and still ship its concentrate across the Pacific. A refinery can run far below its nameplate because it cannot buy feed. A port can be the outlet for a whole copper belt without handling a tonne of its own production. Begin with one asset, read the selected record and decide what the map actually establishes before drawing a wider conclusion.

The atlas brings together the owner's documentary research on copper mines and projects, concentrators and tailings facilities, smelters and acid plants, refineries and SX-EW plants, rod mills and fabricators, ports, rail lines, warehouses and desalination plants, plus the headquarters of the companies that control them. It adds concentrate, refined copper and scrap flows, money items such as royalties and taxes, and a table of sourced facts. Search for a mine, company, plant, port or country; select a result; then switch off the layers that are not relevant to your question.

The reference model is 2025, built into this explorer on 13 November 2026. That does not mean every number is a calendar-2025 observation. Some records use fiscal years ending in June 2025 or March 2026, some use guidance or part-year figures, and some describe earlier years. The page keeps those distinctions visible rather than presenting a live copper market monitor. It does not track prices, exchange stocks or shipments in real time.

Read contained copper, concentrate and ore

Quantities in the atlas are thousand tonnes of contained copper, abbreviated kt. That is the metal inside a product, not the weight of the product. The difference matters at every step. A mine moves hundreds of tonnes of rock for each tonne of copper; the owner's research models about 130 tonnes of ore per tonne of copper at a typical modern grade, before waste rock. Copper concentrate usually carries roughly a quarter of its weight as copper, so a cargo of one million tonnes of concentrate holds about a quarter of a million tonnes of metal. Cathode, the refined product, is about 99.99 percent copper.

Ore grade is a percentage of copper in rock. A grade of 0.6 percent means about six kilograms of copper in a tonne of ore, before losses in the mill. Grade explains why two mines of similar output can have very different footprints, water needs and costs, and why falling grades push output down even when a pit keeps moving the same tonnage.

Company reports use other units too. North American producers publish million pounds, Mlb; one million pounds is about 454 tonnes. Some polymetallic mines publish copper-equivalent output, which converts zinc, lead, gold or silver into copper at assumed prices. The atlas shows such figures as text and does not treat them as copper tonnes.

Follow the layers from pit to wire

Mines appear as copper circles sized by a single-site 2025 copper figure when the record has one. Projects and artisanal areas are open rings; many are years from production. Concentrators and tailings facilities are small diamonds that sit beside the mines they serve.

Smelters are rose squares and refineries and SX-EW plants are teal squares. They are sized by recorded output, or by recorded capacity when no output is given. Rod mills and fabricators are violet triangles. Ports, rail lines and warehouses are blue triangles pointing down, and desalination plants are small blue circles. Company headquarters are switched off by default because a head office is not where copper is produced.

Flows are curves. Orange curves carry concentrate or ore, green curves carry refined copper such as cathode, rod or wire, and dashed brown curves carry scrap, anode, blister or mixed products. Line width follows the recorded copper tonnage where there is one.

Separate mining, smelting and refining

Copper changes hands and countries between the pit and the wire, and each step has its own geography. According to the ICSG figures used by the research, world mine output in 2025 was about 23.2 million tonnes, and about 5.1 million tonnes of that was SX-EW cathode produced at mines by leaching and electrowinning, without a smelter. The rest left mines as concentrate. Smelters turn concentrate into anode, and refineries turn anode into cathode; the ICSG puts world refined output at about 28.7 million tonnes because refineries also treat scrap.

The countries that mine copper are not the countries that refine it. The research's fact-check puts China at roughly 8 percent of world mine output but about 48 percent of refined output. Chile mines nearly a quarter of the world's copper, yet its smelters recorded their lowest output on record in 2025. Peru ships almost everything as concentrate. Japan, South Korea, India and Spain run large smelters on imported feed.

Processing economics shape these maps. Smelters earn treatment and refining charges from miners. The model's facts record a 2026 benchmark charge of zero and negative spot charges, which strain smelters outside China. Sulphuric acid, a by-product of smelting, matters to leach operations in Chile and the Copperbelt, so acid plants and acid prices appear in the records too.

Read flows as documented relationships

A flow record links an origin and a destination named in the research: a mine to its port, a port to a smelter, or a country to a country. Many tonnages are customs totals; others are a mine's output assigned to its known route, or a gross concentrate tonnage converted to copper at an assumed grade, which the research marks as modelled.

The model gives endpoints as words, not coordinates. This explorer places an endpoint at a model site when the wording names one, otherwise at a country hub point taken from the owner's video build. The selected panel says how each end was placed, including where the match was made on name words. Endpoints such as "World", "Asia", "export" or several countries at once are not drawn, so more than a third of the flow records appear only in the table. A drawn line is a documented relationship or trade total between representative points, not a traced shipment.

Customs data records declared movements. Large flows in 2025 reflect real supply chains but also tariff expectations: the research records about 1.63 million tonnes of refined copper imported by the United States, some 700 thousand tonnes more than it used, ahead of a tariff that never came for cathode. Avoid summing flows into a balance; origins, periods and bases differ.

Treat group totals and modelled values carefully

Copper companies report in many ways: by mine, by division, by group, at 100 percent or attributable share. The research keeps that wording. A record can hold a figure for a whole district, a combined pair of mines, a company's cathode from all its plants, or a country total. This explorer reads a number only when the text gives the record's own copper. District and combined figures size the record that holds them, and the linked records that say "included above" carry no number of their own. Group, company and country totals are shown in full but do not set symbol size or enter sums.

Values the research modelled are its own estimates: a share of a group total, a part-year figure scaled to a year, or a conversion from pounds or gross concentrate. Many model money values are conversions at the 2025 LME average of US$9,939 a tonne. The selected panel spells out each modelled value in words and keeps the method the research gave.

Separate reserves, resources and output

Output is what was produced in a period. Reserves are the part of a deposit that can be mined at a profit under stated price, cost and permitting assumptions, reported as proved and probable. Resources are larger and less certain; resources are not reserves. Few site records in this model carry a reserve figure, so the atlas does not map reserves.

The USGS figure used by the research puts world reserves at about 980 million tonnes of copper. Dividing that by annual mine output gives a static ratio of about four decades, which is often misread as the time until copper runs out. Reserves rise with price, technology and drilling, so the ratio describes a working inventory, not the copper left in the ground. The research also notes that it read the USGS country table from pages that reproduce it, because the original PDF could not be opened.

Keep the period, basis and verification beside the number

Only 94 of the 250 mine records carry a single-site 2025 copper figure. Together they sum to about 14.3 million tonnes, against roughly 23 million tonnes of world mine output in the ICSG and USGS figures. The gap is coverage, not a correction: many mines carry no figure, are included in a district or group total, or are not in the model at all. Treat the sum as an indicator of how much of the world the named records describe, never as a total.

Periods differ. BHP and Indian producers report fiscal years, Mexican records cite 2024 national statistics, and some records use guidance or a part-year figure. Stages differ too: one record measures copper in concentrate, another cathode, another anode. The selected panel names the period and stage whenever the text gives them.

The owner's research flags many fields for checking, most often notes, owner stakes and start years, and some records cite background knowledge that was not verified. A flag is not a correction; it tells you which fields need a primary source before they carry weight in a decision.

Use four starting points to test a story

Start with Escondida to read the world's largest copper mine: one record with output, capacity, owners, process route and water source, which the research says out-produced all of Codelco's own divisions in 2025. Then open the Guixi smelter to see how a Chinese plant's smelter and refinery capacities are recorded and which figure sizes the symbol.

Use the Lobito corridor to read a port whose wording describes cargo and a target, not copper tonnes, and follow the rail and cathode flows that run to it from the Copperbelt. Then select Kamoa-Kakula to see how a whole-complex figure, linked mine records and a new smelter sit together.

These starting points follow the documentary's tonne from a pit, through a smelter and a corridor, to the coast. They do not establish that a particular cargo made that journey. When an argument needs that connection, add contract, customs or company evidence appropriate to the claim.

Build a presentation around inspectable evidence

Choose one decision question before exporting a map image: concentration of mine supply, dependence on Chinese smelting, exposure to a single corridor or port, or the effect of a smelter closure on a region. Use the map for location and relationships, then a small table for quantities. Keep the unit, period, stage and verification status next to the conclusion.

Avoid a crowded slide that asks a reader to infer the whole model. Show the relevant region or layer, state the observation in the headline and keep the source limitation in a legible note. The consulting presentation guide can help structure the argument; the global oil atlas, global gas atlas, global coal atlas and global gold atlas provide companion views of other commodity systems.

Retain attribution when reusing the downloadable snapshot. The JSON contains the rendered atlas, the parsed readings and the provenance notes, not every research file in the source project, such as the price, stock, control and demand tables. Review publisher terms and verify the inputs central to your conclusion.

COMMON QUESTIONS

Before drawing a conclusion.

Is this a live copper map?

No. It is a documentary research snapshot built into this explorer on 13 November 2026, with a 2025 reference model and mixed calendar, fiscal and part-year input periods. It does not track live prices, exchange stocks, mine output or shipments.

Why do the mine figures not add up to world output?

Only 94 of 250 mine records carry a single-site 2025 copper figure. They sum to 14.3 Mt, against about 23.2 Mt of world mine output in ICSG figures and about 23.0 Mt from the USGS. Many mines have no figure, sit inside a district or group total, or are not in the model.

Why is a well-known mine or plant drawn small?

A record is sized only by its own copper figure. When the research gives a group, company or country total, an ore or cargo tonnage, a copper-equivalent figure or says the output is included in another record, the wording is shown in full but the symbol keeps the smallest size.

Does a flow line prove where a cargo went?

No. Flow endpoints are placed at a named model site or at a country hub point, and many tonnages are customs totals or modelled conversions. A line is a documented relationship or trade total, not a traced shipment. Flows to “World”, “Asia” or unnamed buyers are listed but not drawn.

What does the verification flag mean?

The owner’s research marked that field for checking, or cited background knowledge that was not verified. The selected panel names the flagged fields. A flag is not a correction; check a primary source before relying on that record.

SOURCES & ATTRIBUTION

A visible evidence trail.

Built from the owner’s copper documentary research. Most records also cite their own company filing, ministry, customs or press source in the selected panel. Dates below are retained project labels, not fresh verification of every mine, plant, flow or payment.

Company filings and results

Mine, smelter and refinery output, capacities and ownership from annual reports, 10-K and 20-F filings and production reports (Freeport-McMoRan, BHP, Codelco, Southern Copper, Ivanhoe, Glencore, Jiangxi Copper, Aurubis, KGHM and others). Each record carries its own source.

Publisher’s usage terms apply ↗

XLSlides built this explorer from the owner’s copper model files, reusing the design of the gold atlas. Values and source strings are kept as written; research markers are spelled out; quantities are parsed only where the text states a copper tonnage for the record itself. Flow endpoints were placed by matching site names or country hubs, following the owner’s video build. No fresh asset-by-asset validation, live tracking or publisher endorsement is claimed. Download attribution notes.

FROM AN ASSET TO AN ARGUMENT

Make the copper story clear.

Build a focused, sourced presentation that keeps reporting periods, product stages, modelled estimates and verification status visible.

Build a research presentation