The IEA’s 2026 assessment projects a 25% gap against primary copper supply requirements in 2035 under its stated-policy comparison, down from around 30% in the preceding outlook. That is an assessment of existing and announced mining projects against a modeled requirement, not evidence that a quarter of today’s orders cannot be filled. This reference brings the production tables, US historical inputs, source revisions, recycling boundaries and forecast definitions together. All quantities retain their source years and estimate status. (IEA, Global Critical Minerals Outlook 2026; accessed, Nov 2026).
Key Takeaways
- 23 million metric tons was estimated world mine output in 2025. (USGS 2026, world table; accessed, Nov 2026).
- 29 million metric tons was estimated world refinery output in 2025, a different production stage. (USGS 2026, world table; accessed, Nov 2026).
- Chile mined 5.3 million tons, about 23.04% of the rounded world total by AMR calculation. (USGS 2026; accessed, Nov 2026).
- China refined 14 million tons while mining 1.8 million tons in the same source year. (USGS 2026; accessed, Nov 2026).
- US mine output was 1 million tons, compared with 850,000 tons of refinery production in 2025 estimates. (USGS 2026; accessed, Nov 2026).
- US net import reliance reached an estimated 57% of apparent consumption in 2025. (USGS 2026; accessed, Nov 2026).
- US 2024 mine output was revised from 1.1 to 1.05 million tons between the two MCS editions. (USGS 2025; accessed, Nov 2026). (USGS 2026; accessed, Nov 2026).
- 160,000 tons of old scrap and 760,000 tons of new scrap were recovered in the 2025 US estimates. (USGS 2026; accessed, Nov 2026).
- 60,000 tons of secondary refinery output is a stage measure that overlaps the scrap stream. (USGS 2026; accessed, Nov 2026).
- 42% of US copper product use was building construction, using the end-use classification cited by USGS. (USGS 2026; accessed, Nov 2026).
- The projected 2035 gap narrowed by about 5 percentage points, comparing 30% with 25%. (IEA 2026; accessed, Nov 2026).
- Reserves are a stock, production is a flow. Dividing the two does not establish an exhaustion year. (USGS 2026, reserves and resource notes; accessed, Nov 2026).
Which countries produce the most copper?
Chile is the largest named mine producer in the USGS 2025 estimates. Congo (Kinshasa), Peru and China follow. AMR ranks the named producing countries in the source table and retains “Other countries” in the complete world comparison rather than presenting the list as a census of every country. (USGS 2026, world table; accessed, Nov 2026).
View chart data
2025 estimates, thousand metric tons of copper; selected named producers.
| Country | Mine output (kt) |
|---|---|
| Chile | 5300 |
| Congo (Kinshasa) | 3200 |
| Peru | 2700 |
| China | 1800 |
| Russia | 1300 |
| United States | 1000 |
Values and definitions are also readable in HTML. Source precision and scope apply.
| Country | Rank | Mine output estimate (kt) | Share of rounded world total (%) | Change from 2024 (%) |
|---|---|---|---|---|
| Chile | 1 | 5300 | 23.0435 | -3.8113 |
| Congo (Kinshasa) | 2 | 3200 | 13.913 | 7.0234 |
| Peru | 3 | 2700 | 11.7391 | -1.4599 |
| China | 4 | 1800 | 7.8261 | -2.1739 |
| Russia | 5 | 1300 | 5.6522 | 27.451 |
| United States | 6 | 1000 | 4.3478 | -4.7619 |
| Zambia | 7 | 940 | 4.087 | 14.2163 |
| Australia | 8 | 730 | 3.1739 | -4.5752 |
| Indonesia | 9 | 710 | 3.087 | -29.703 |
| Kazakhstan | 10 | 710 | 3.087 | -1.9337 |
| Mexico | 11 | 690 | 3.0 | -3.7657 |
| Canada | 12 | 500 | 2.1739 | -2.9126 |
| Poland | 13 | 410 | 1.7826 | 2.5 |
| India | 14 | 23 | 0.1 | -14.8148 |
What exactly does the ranking measure?
The unit is copper content in mined material, expressed in thousand metric tons in the source. It is not the mass of all rock moved, concentrate shipped or copper ore processed. One thousand source units equals one million metric tons of contained copper. Chile’s source value of 5,300 therefore becomes 5.3 million tons, without implying a mine moved only that much total material.
How are world shares calculated?
AMR divides each named country’s 2025 estimate by the published rounded world total of 23,000 thousand tons. Chile’s calculation is 5,300 / 23,000 × 100 = 23.0435%. The source inputs have limited precision; the calculation’s extra decimal places support reproducibility, not a claim that production is known to four decimal places. The named countries and residual may not reconcile exactly to a rounded world total. Preserve that rounding rather than distribute a manufactured balancing adjustment. (USGS 2026; accessed, Nov 2026).
Can production rank tell us where copper will be available?
Rank identifies origin at the mining stage. It does not establish ownership, the destination of concentrates, processing capacity, contracted supply or which country controls a company. A disruption at a producer can matter to global feedstock even when its domestic refinery output is small. Use the stage comparison below before describing a production leader as a finished-copper supplier.
Why do mining and refining rankings differ?
The world table estimates China’s 2025 refinery output at 14,000 thousand tons and mine output at 1,800 thousand tons. AMR’s arithmetic difference is 12,200 thousand tons, but it is not a net-import estimate. The measures refer to different stages and inputs. (USGS 2026, world table; accessed, Nov 2026).
View chart data
2025 estimates; mine and refinery output are separate stages, not a trade balance.
| Country | Mine (kt) | Refinery (kt) |
|---|---|---|
| China | 1800 | 14000 |
| Chile | 5300 | 1700 |
| United States | 1000 | 850 |
Values and definitions are also readable in HTML. Source precision and scope apply.
| Country | Mine 2025 estimate (kt) | Refinery 2025 estimate (kt) | Refinery minus mine (kt) | Interpretation |
|---|---|---|---|---|
| United States | 1000 | 850 | -150 | Different stages; difference is not net imports |
| Australia | 730 | 460 | -270 | Different stages; difference is not net imports |
| Canada | 500 | 320 | -180 | Different stages; difference is not net imports |
| Chile | 5300 | 1700 | -3600 | Different stages; difference is not net imports |
| China | 1800 | 14000 | 12200 | Different stages; difference is not net imports |
| Congo (Kinshasa) | 3200 | 2800 | -400 | Different stages; difference is not net imports |
| Germany | 0 | 610 | 610 | Different stages; difference is not net imports |
| India | 23 | 620 | 597 | Different stages; difference is not net imports |
| Indonesia | 710 | 400 | -310 | Different stages; difference is not net imports |
| Japan | 0 | 1400 | 1400 | Different stages; difference is not net imports |
| Kazakhstan | 710 | 500 | -210 | Different stages; difference is not net imports |
| Korea, Republic of | 0 | 610 | 610 | Different stages; difference is not net imports |
| Mexico | 690 | 480 | -210 | Different stages; difference is not net imports |
| Peru | 2700 | 340 | -2360 | Different stages; difference is not net imports |
| Poland | 410 | 560 | 150 | Different stages; difference is not net imports |
| Russia | 1300 | 950 | -350 | Different stages; difference is not net imports |
| Zambia | 940 | 270 | -670 | Different stages; difference is not net imports |
| Other countries | 3000 | 2100 | -900 | Different stages; difference is not net imports |
Why can a refinery country produce more than it mines?
Refineries can process imported feedstock and secondary material. Annual production also crosses inventory and timing boundaries. A subtraction between refinery and mine output shows the scale difference between stages; it cannot identify the amount of imported concentrate, metal recovered from scrap or stock changes without those separate inputs.
Which world table should a writer use?
Use mine output for the origin of newly extracted copper. Use refinery output for the production of refined copper. Use import data for trade dependence. The complete USGS table below keeps both production stages and reserves visible so a reader can choose the relevant measure rather than quote the largest number available.
| Geography | Mine 2024 (kt) | Mine 2025 estimate (kt) | Refinery 2024 (kt) | Refinery 2025 estimate (kt) | Reserves (kt) |
|---|---|---|---|---|---|
| United States | 1050 | 1000 | 921 | 850 | 47000 |
| Australia | 765 | 730 | 434 | 460 | 100000 |
| Canada | 515 | 500 | 324 | 320 | 7000 |
| Chile | 5510 | 5300 | 1940 | 1700 | 180000 |
| China | 1840 | 1800 | 12400 | 14000 | 41000 |
| Congo (Kinshasa) | 2990 | 3200 | 2560 | 2800 | 80000 |
| Germany | 0 | 0 | 597 | 610 | 0 |
| India | 27 | 23 | 545 | 620 | 2200 |
| Indonesia | 1010 | 710 | 349 | 400 | 21000 |
| Japan | 0 | 0 | 1570 | 1400 | 0 |
| Kazakhstan | 724 | 710 | 498 | 500 | 20000 |
| Korea, Republic of | 0 | 0 | 604 | 610 | 0 |
| Mexico | 717 | 690 | 489 | 480 | 53000 |
| Peru | 2740 | 2700 | 385 | 340 | 85000 |
| Poland | 400 | 410 | 589 | 560 | 33000 |
| Russia | 1020 | 1300 | 896 | 950 | 80000 |
| Zambia | 823 | 940 | 189 | 270 | 21000 |
| Other countries | 2850 | 3000 | 2310 | 2100 | 210000 |
| World total (rounded) | 23000 | 23000 | 27600 | 29000 | 980000 |
USGS explicitly defines the dash in this world table as zero. Those cells can therefore be transcribed as zero with that source convention recorded. The Australia reserve cell includes footnote 7 before 100,000; the footnote is not part of the number. AMR retains 100,000 thousand tons, not 7,100,000. These are transcription issues with large consequences for reuse. (USGS 2026, table footnotes; accessed, Nov 2026).
Are reserves the same as resources?
No. The source distinguishes reported reserves from a broader resource assessment, including identified and undiscovered material. A reserve estimate is conditional on knowledge, economics and the source’s classification at the time. It is neither a fixed physical limit nor an immediately deliverable inventory.
What does the United States produce and consume?
USGS estimates 2025 recoverable mine production at 1,000 thousand tons, primary refinery production at 790 thousand tons and secondary refinery production at 60 thousand tons. Primary plus secondary refinery output gives the 850 thousand tons in the world table. (USGS 2026, salient statistics; accessed, Nov 2026).
View chart data
2025 US estimates, thousand metric tons; apparent and reported consumption differ.
| Measure | Value |
|---|---|
| Mine output | 1,000 kt |
| Reported refined consumption | 1,700 kt |
| Apparent consumption | 2,200 kt |
| Net import reliance | 57% of apparent consumption |
Values and definitions are also readable in HTML. Source precision and scope apply.
| Measure | 2021 | 2022 | 2023 | 2024 | 2025 estimate | Unit |
|---|---|---|---|---|---|---|
| Mine, recoverable | 1230 | 1230 | 1130 | 1050 | 1000 | thousand metric tons copper |
| Primary refinery output | 931 | 917 | 843 | 882 | 790 | thousand metric tons copper |
| Secondary refinery output | 49 | 40 | 39 | 39 | 60 | thousand metric tons copper |
| Copper from old scrap | 169 | 152 | 137 | 140 | 160 | thousand metric tons copper |
| Refined imports | 919 | 732 | 771 | 903 | 1700 | thousand metric tons copper |
| Refined exports | 48 | 27 | 29 | 72 | 110 | thousand metric tons copper |
| Reported refined consumption | 1750 | 1720 | 1580 | 1580 | 1700 | thousand metric tons copper |
| Apparent consumption | 1970 | 1810 | 1680 | 1860 | 2200 | thousand metric tons copper |
| Yearend refined stocks | 117 | 84 | 128 | 123 | 450 | thousand metric tons copper |
| Net import reliance (%) | 44 | 41 | 42 | 45 | 57 | percent |
Why are there two consumption measures?
Reported refined consumption describes the reporting category in the salient table. Apparent consumption is an accounting measure incorporating primary refined production, old scrap, refined imports, refined exports and stock adjustments. The 2025 values are 1,700 and 2,200 thousand tons respectively. They should not be interchanged as denominators for the same percentage.
What does 57% import reliance mean?
The source defines net import reliance using refined imports minus refined exports, adjusted for refined stocks, as a percentage of apparent consumption. It is not simply imports divided by refinery output. Refined imports are estimated at 1,700 thousand tons and refined exports at 110 thousand tons, but the stock adjustment means their difference alone is not the complete numerator. The 57% figure should remain attributed to USGS rather than reconstructed from incompatible rounded components.
What does the US history actually show?
The current source edition gives US mine output of 1,230 thousand tons in 2021 and 1,000 thousand tons in 2025. AMR calculates a decrease of about 18.70% across those endpoints. This is a comparison within one edition. It is different from taking an old estimate for 2024 out of another edition and labeling the difference a one-year production change. Stocks, trade and refinery figures can move independently, so a mine-output change alone cannot explain the full import-reliance outcome.
How much does recycling contribute to copper supply?
USGS reports estimated recovery of 160 thousand tons from old scrap and 760 thousand tons from new scrap in 2025. Their combined 920 thousand tons includes recovery into refined metal, alloys and other forms. It is not equivalent to the 60 thousand tons reported as secondary refinery production. (USGS 2026, recycling paragraph and notes; accessed, Nov 2026).
View chart data
2025 US estimates; these categories must not all be added together.
| Measure | Amount and boundary |
|---|---|
| Old scrap | 160 kt, post-consumer |
| New scrap | 760 kt, manufacturing |
| Secondary refinery output | 60 kt, overlapping stage |
Values and definitions are also readable in HTML. Source precision and scope apply.
| Measure | 2025 estimate (kt) | Scope |
|---|---|---|
| Old scrap | 160 | Post-consumer material, all reported recovery forms |
| New scrap | 760 | Manufacturing scrap |
| Secondary refinery output | 60 | A production stage; overlaps scrap recovery, do not add again |
What is old scrap versus new scrap?
Old scrap is post-consumer material returning after use. New scrap arises during manufacturing and fabrication. Their different timing matters: new scrap can return quickly within production, while old scrap depends on the life of products already in use. A high new-scrap recovery figure does not prove an equivalent rate of end-of-life consumer collection.
Why not add all three numbers?
Secondary refinery output measures a processing stage. Material entering that stage can already be counted in the scrap recovery streams. Adding 160, 760 and 60 would mix feedstock recovery and output stages, creating a double count. The source’s statement that scrap contributes about 30% of US supply uses a broader supply definition, including new scrap, than the apparent-consumption line used for net import reliance. One cannot use the percentage from one definition with the denominator from the other.
Does recycling remove the need for mines?
Recycling returns existing material, while mine supply introduces newly extracted copper. The two are related but distinct. Whether secondary supply meets a particular requirement depends on collection, product lifetimes, process losses, quality and the demand definition. The available source supports the measured scrap quantities and its own approximate supply share; it does not establish a universal percentage of future mining that can be avoided.
Is a copper shortage forecast for 2035?
The IEA’s 2026 Outlook compares expected mine supply from existing and announced projects with primary supply requirements. The stated-policy comparison shows a projected gap around 25% in 2035, narrower than around 30% in the previous assessment. AMR records both as scenario-specific projections, with the prior figure identified as reported retrospectively by the 2026 source. (IEA 2026, outlook and chart notes; accessed, Nov 2026).
View chart data
2035 primary copper requirement, STEPS; rounded projections described in IEA 2026.
| Assessment | Gap | Implied coverage |
|---|---|---|
| Prior assessment | around 30% | around 70% |
| 2026 assessment | 25% | 75% |
| AMR difference | about 5 percentage points narrower | not a current shortage |
Values and definitions are also readable in HTML. Source precision and scope apply.
| Assessment | Target year | Scenario | Projected gap (%) | Implied supply coverage (%) | Evidence |
|---|---|---|---|---|---|
| Prior Outlook as reported by IEA 2026 | 2035 | STEPS | 30 | 70 | Approximate prior projection, not actual shortage |
| IEA 2026 Outlook | 2035 | STEPS | 25 | 75 | Announced-project base case; not an observed shortage |
What is the denominator of the gap?
Primary supply requirements are total demand net of secondary supply, with an allowance for refining losses. They are not total end-user demand without adjustment. Expected supply is mined or raw material output associated with the source’s project pipeline. A gap against primary requirements therefore cannot be quoted as a 25% shortfall of all copper demand without changing the meaning.
How large was the revision?
Moving from about 30% to 25% narrows the gap by about 5 percentage points. Relative to the earlier gap, that is a reduction of about 16.67%, not 5%. The implied coverage of the requirement changes from about 70% to about 75%. Those are arithmetic restatements of rounded source percentages, not separately measured quantities. The absolute tonnage cannot be inferred without the matching scenario requirement.
What could change the result?
Project delivery, output at existing mines, secondary supply, losses and demand assumptions all affect the balance. The chart’s current-policy, stated-policy and high-demand scenarios answer different conditional questions. New projects or revised expectations can change a later assessment even when the target year remains 2035. Preserve scenario, edition and target year in an attribution; a bare claim that “copper will run out” is unsupported by this comparison.
Why do two sources report different values for the same year?
A value can change because an estimated year is revised, not because a different physical year occurred. The 2025 MCS estimated 2024 US mine output at 1,100 thousand tons; MCS 2026 records 1,050 thousand tons for that same year. AMR calculates a revision of minus 50 thousand tons, or about minus 4.55% relative to the earlier estimate. (USGS MCS 2025; accessed, Nov 2026). (USGS MCS 2026; accessed, Nov 2026).
View chart data
2024 US mine production, thousand metric tons; source editions differ.
| Source edition | 2024 estimate |
|---|---|
| MCS 2025 | 1100 |
| MCS 2026 | 1050 |
Values and definitions are also readable in HTML. Source precision and scope apply.
| 2024 measure | MCS 2025 estimate | MCS 2026 value | Revision | Unit |
|---|---|---|---|---|
| Mine, recoverable | 1100 | 1050 | -50 | kt |
| Primary refinery output | 850 | 882 | 32 | kt |
| Secondary refinery output | 40 | 39 | -1 | kt |
| Copper from old scrap | 150 | 140 | -10 | kt |
| Refined imports | 810 | 903 | 93 | kt |
| Refined exports | 60 | 72 | 12 | kt |
| Reported refined consumption | 1600 | 1580 | -20 | kt |
| Apparent consumption | 1800 | 1860 | 60 | kt |
| Yearend refined stocks | 70 | 123 | 53 | kt |
| Net import reliance (%) | 45 | 45 | 0 | percentage points |
How should revisions be handled in a trend?
Choose a consistent edition for a historical trend. If the analytical purpose is to study revisions, hold the year constant and compare editions instead. AMR’s revision table does the latter. It includes stock, trade, production and consumption changes so a reused figure can be traced to its source vintage.
What are the main end uses?
| Use | Share (%) |
|---|---|
| Building construction | 42 |
| Electrical and electronic products | 23 |
| Transportation equipment | 18 |
| Consumer and general products | 10 |
| Industrial machinery and equipment | 7 |
The end-use percentages cited by USGS describe copper and copper-alloy products in the stated categories. Building construction accounts for 42%, electrical and electronic products 23%, transportation equipment 18%, consumer and general products 10%, and industrial machinery and equipment 7%. These sum to 100% within that classification. They do not identify a standalone AI, EV or data-center share. Those technologies can span more than one category, and assigning a new percentage would require additional evidence.
Which figure belongs in a headline?
For a production headline, use the measured stage, year and estimate status. For a demand headline, identify consumption type. For a shortage headline, preserve forecast scenario and target year. For a recycling headline, identify whether the metric is scrap recovery, secondary refinery output or a share of supply. That specificity makes the statistic reusable without turning a valid source number into a different claim.
How should this research be quoted?
Use the date and denominator beside the number. Attribute original measurements to their producers and AMR’s calculations or comparisons to American Mineral Registry. The specific finding controls below preserve those qualifications and link the exact evidence. A source-derived estimate is not a transaction price, a current operational reading or a forecast guarantee.
Frequently asked questions
Does 23 million tons mean all the ore mined?
No. USGS reports contained copper. Total ore and waste-rock mass is a different measure.
Why is refinery output greater than mine output?
Refinery output and mine output measure different stages. Secondary feedstock and inventory or trade timing prevent a simple equality.
Which country mines the most copper?
Chile is the largest named producer in the 2025 USGS estimates at 5.3 million metric tons. (USGS 2026; accessed, Nov 2026).
Is China the largest miner?
China is the largest refinery producer in this table, while Chile is the largest mine producer. The distinction changes the answer.
Are the 2025 figures final?
The source marks the 2025 values as estimates. Preserve that label even when the arithmetic is exact.
Does the 25% gap mean a shortage today?
No. It is a conditional comparison for 2035 under an IEA scenario and project pipeline.
Can reserves divided by production predict depletion?
No. That ratio assumes a fixed reserve estimate and fixed production while ignoring future discovery, economics, reclassification, recycling and demand changes.
Can I download the full comparison?
Yes. The data section includes every table, a source-cell ledger, definitions and a calculation bundle without an account or form.
Methodology and Sources
AMR transcribed the complete USGS world copper table and selected salient US statistics from MCS 2026, retaining its thousand-metric-ton copper-content unit, 2025 estimate status, zero convention and reserve footnotes. The named-country mine ranking excludes the residual “Other countries” bucket and zero mine producers; the complete source comparison remains available. Shares divide by the rounded 23,000 thousand-ton world total. US history uses one source edition; the revision table separately compares 2024 values in MCS 2025 and 2026. Percentage changes use (new / old minus 1) × 100; percentage-point changes subtract percentages directly. The forecast table uses the IEA 2026 description of its current and prior assessments and does not claim an independent extraction of absolute scenario tonnage. These are AMR compilations and calculations from producer evidence, not a survey or proprietary market-price database. (USGS 2026; accessed, Nov 2026). (USGS 2025; accessed, Nov 2026). (IEA 2026; accessed, Nov 2026).
The source-cell ledger identifies the underlying table or paragraph for each input. Calculations retain more decimal places in CSV than the source precision warrants for prose; users should quote rounded values with qualifications. Source revisions can change the historical record. The reproduction bundle contains the published CSV inputs, formulas, a validation program and methods; original producer documents remain linked at their official locations.
Glossary
- Mine production. Copper content recovered at the mining stage.
- Refinery production. Output from refining, including primary and secondary production.
- Old scrap. Post-consumer material recovered after use.
- New scrap. Manufacturing and fabrication scrap.
- Apparent consumption. A source-defined accounting measure using production, scrap, trade and stock changes.
- Net import reliance. Source-defined net imports with stock adjustments relative to apparent consumption.
- Reserves. A source-classified stock under the conditions of its assessment.
- Primary requirement. Demand net of secondary supply, with source adjustments for losses.
- STEPS. The IEA Stated Policies Scenario.
- Vintage. The particular edition of a source or model assessment.
Sources
- USGS: Copper, MCS 2026. World production, reserves, US history, recycling and footnotes.
- USGS: Copper, MCS 2025. Earlier 2024 estimates for the revision comparison.
- IEA: Global Critical Minerals Outlook 2026, Outlook. Scenario definitions and projected gap comparison.