Copper in the Age of AI analyzes the global outlook for copper supply and demand through 2040, focusing on copper’s essential role in meeting the growing requirements of electrification, digitalization, and technologies such as AI, data centers, electric vehicles, and defense. It represents the collaboration of groups across S&P Global. The analysis and metrics developed during the course of this research represent the independent analysis and views of S&P Global. The study makes no policy recommendations.

The cross enterprise S&P Global teams involved in the study include:

S&P Global Energy

• Critical Minerals and Energy Transition Consulting provides strategic analysis and advice and quantifiable, data-driven actionable insights on the metals and mining sector for decision makers in both companies and governments.
• Metals and Mining Research provides comprehensive and proprietary datasets, analysis and forecasts, spanning the full supply chain of the upstream mining sector. A global team of analysts deliver monthly, quarterly, annual and ad hoc research that covers capital market activity, exploration trends, corporate strategies, M&A activity, production cost modeling, supply and demand fundamentals, and industrial and battery metals prices.
• Electric Power and Renewable Research analyzes global electricity markets and policy developments to provide strategic insights and data-based forecasts for generators and distributors, policymakers and regulators, investors and lenders, and major consumers, including industrial and tech companies.

S&P Global Market Intelligence

• Global Insight, built on a foundation of data and technical expertise, creates customized client solutions via market analysis, models, scenarios, indices, and early warning systems that inform decision-making and strategy across a range of workflows and sectors, including economic and industry planning, geopolitics and risk intelligence, and supply chain and logistics.
• 451 Research provides essential insight into key trends driving digital transformation across the technology landscape. Critical for this study is 451’s Data Center Knowledge Base, which provides detailed intelligence on over 14,000 data center facilities in more than 135 countries and 2,700 owners/providers.

Ever since Thomas Edison’s enterprise laid 80,000 feet of copper wires under streets in Lower Manhattan in 1882, lighting up one square mile, copper has proved its mettle as the metal of electrification. In the century and a half since then, as copper has gone on to wire the world, the staggering growth in consumption has turned it into one of the most important materials of modern civilization. But, without significant adjustments, copper supply faces a growing challenge of keeping up with the accelerating pace of electrification.

The importance of copper has been underlined over the last half decade as a number of countries have deemed it a “critical metal”, including, in 2025, the United States (US). And with good reason. Copper is the connective artery linking physical machinery, digital intelligence, mobility, infrastructure, communication, and security systems. All of this has made the future availability of the metal a matter of strategic importance. The United States’ designation of copper as a critical mineral underlines its essential role in enabling the infrastructure, technologies, and security systems that will shape the coming decades.

S&P Global's comprehensive study identifies a transformative trajectory for copper demand, projecting a surge from 28 million metric tons in 2025 to 42 million metric tons by 2040 – a 50% increase that underscores the metal's pivotal role in multiple technological and economic domains. However, meeting the call on copper confronts significant supply obstacles both above and below ground. The study projects a potential 10 million metric ton copper shortfall by 2040 without meaningful supply expansion. This demand growth – and addressing the looming challenges to meeting it – is what this report is about.

But what is driving this demand? It arises from the fact that copper is essential for the generation, transmission, and use of electricity. But the demand for copper will outrun supply unless there is major adjustment across the copper supply system.

Here, in short, is the quandary: copper is the enabler of electrification, but the accelerating pace of electrification is an increasing challenge for the metal.

In S&P Global’s base case, global electricity demand will increase by nearly 50% by 2040. And this surging electrification is advancing around the world. In the United States, for a quarter century, electricity consumption hardly rose year over year, but it is now beginning to grow at what could be 2.5% annually. In China – with an electricity market more than double that of the United States – it will grow at 3.2% per year between now and 2040. In India, it will be 4.2% per year.

The demand for copper is growing along four vectors, each of which adds to the pyramiding call on copper. The newest to emerge is one that was not even evident four years ago and yet today is highly visible in terms of global transformation for both work and life. That, of course, is artificial intelligence (AI). While AI has long been in development, it only “broke through” in November 2022, with the debut of ChatGPT. That launched the “AI Race”, which runs on electricity. In 2025, half of US Gross Domestic Product (GDP) growth is attributed to AI spending – largely on computer chips, data centers, and the electric power systems on which they run.

This explosive growth of AI and data centers has introduced a new, rapidly expanding vector of copper demand. Data centers are electricity-intensive, and their proliferation is driving massive investments in both direct copper use (for power delivery, cooling, and IT infrastructure) and in the electric grid infrastructure that supports them. By 2030, data centers alone could rise from today’s 5% to 14% of US electricity demand, with copper a critical enabler all along the way. What is still to play out is the indirect impact of AI in terms of the electric infrastructure needed to meet the enormous demands of users – and the impact that AI will have in generating industrial, commercial, creative, and personal applications that will lead to further cycles of copper demand.

While AI is creating a new vector of copper demand, it is not the largest by any means. But the reason that we call this paper “Copper in the Age of AI” is because the requirements of AI underline the essential and foundational role of expanded electricity supply – and thus the need for more copper.

It is “core economic demand” that we cite as the first vector – from appliances and computers to construction and manufacturing – going back to when Thomas Edison’s light bulb candles and kerosene lamps. And this vector of demand – the largest – continues to grow. In the developing world, a combination of urbanization, rising incomes, and changing building practices means electricity use and thus more copper. One vivid example: the developing world is projected to add as many as two billion new air conditioners by 2040. In the United States, the reshoring of manufacturing and the resulting growth in electricity consumption is driving electric utilities to add more generation, more transformers, and more transmission and distribution lines.

A second vector of copper demand has only emerged over the last decade – “energy transition and addition.” Electric vehicles (EVs) require 2.9 times more copper than a conventional car, and the population of EVs is growing. The number of electric cars sold worldwide in 2025 was 25% greater than total new cars sold in the United States, the world’s second largest new car market. Solar and wind require a lot of copper, and over 90% of the new electric generating capacity installed in 2025 worldwide was solar and wind. Another new demand for copper is for the batteries being deployed to store renewably generated electricity. Transmission and distribution systems are being expanded worldwide.

But energy transition takes another form as well – it is also populations in the developing world moving from wood and waste for their heating and cooking to commercial energy, including electricity. Africa is home to almost 20% of the world’s population but is grossly underserved in terms of electricity. Copper will be integral to the systems that are rolled out to meet the need for electricity across that continent.

The final vector of current copper demand is defense. Rising geopolitical tensions and the electrification of military systems and the battlefield itself are driving up spending on defense and the push for new technologies. The investment in these technologies and systems is “inelastic” – given the national security stakes. Notable is the pledge by NATO members to increase defense spending to five percent of GDP. Modern weaponry, communications, and infrastructure are increasingly copper-intensive, and defense-driven demand is projected to triple by 2040.

And now a possible new vector of demand is on the horizon – humanoid robots. There is much variance in projections for their scale by 2040 – varying from tens of millions to hundreds of millions to a billion or more. Whatever the actual number, these humanoids will not just be wired – but heavily wired – with copper.

Even as global demand is accelerating along these vectors, current supply is on course to decline as existing resources age. Without meaningful expansion of supply, the result could be a 10 million metric ton shortfall by 2040.

Figure ES- 1. Total copper market balance (2020–2040)
MMt Cu

Meeting the growing demand is fraught with challenges both above and below ground. The supply response is multi-faceted but constrained:

• Mined copper – “primary supply”: Mining faces declining ore grades, rising costs, and increasingly complex extraction conditions. Without significant new investment, output from existing mines will decline, and the pipeline of new projects is hampered by long development timelines – averaging 17 years – resulting from permitting delays, and above-ground risks such as regulatory uncertainty, environmental activist opposition, shifting government terms, and rising costs.
• Recycling – “secondary supply”: While copper’s recyclability offers additional supply, such secondary supply alone cannot close the gap. Even with very aggressive improvements in collection and processing, recycling could provide at best only about a third of total supply by 2040, leaving a substantial shortfall.
• Processing criticality: Smelting and refining capacity, especially concentrated in China, have become critical nodes in the supply chain. The economics of processing are increasingly precarious, with treatment and refining charges under pressure and with regional disparities in operating costs and regulatory environments. The geographic concentration – estimated at 40% to 50% of total capacity – amplifies systemic risks and exposes the supply chain to geopolitical shocks.

What distinguishes copper from other metals is its exceptional conductivity of electricity – exceeded only by a precious metal, silver. This conductivity, along with the metal’s durability and recyclability, makes substitution difficult in most applications. While aluminum, plastics, and fiber optics compete in select uses, copper remains the preferred and/or essential material for safety, performance, heat management, and sustainability. Substitution, miniaturization and “thrifting” (using less copper per application) are limited by technical and economic factors, and the bulk of feasible substitutions are considered to have already occurred. The price ratio of copper to aluminum remains elevated, but further displacement may be limited without major technological breakthroughs.

Governments are increasingly recognizing the strategic importance of stable and competitive mineral supply chains. Emerging modes of international cooperation and the growing role of sovereign wealth funds are offering new approaches to secure and diversify critical mineral supplies.

This report underscores the urgency of policy action, investment, and innovation across the copper value chain. Meeting rising demand in the coming decades will require considerable effort and innovation across the entire value chain:

• Accelerating mine development: Streamlining permitting and rationalizing judicial reviews, fostering stable investment frameworks, and leveraging new technologies are essential to shortening development timelines and unlocking new supply.
• Expanding processing capacity: Diversifying smelting and refining capacity beyond current hubs, incentivizing innovation, and aligning tariffs and industrial policy are critical to reducing systemic vulnerabilities – but also come with challenging economics and additional costs as well as regulatory obstacles.
• Enhancing recycling: Investment in collection infrastructure, regulatory incentives, and international cooperation can boost recycling rates, but cannot replace primary supply growth.
• Talent and skills: Addressing the talent gap in mining and processing is vital, as the industry faces a wave of retirements and declining enrollment in technical programs.

Copper’s role as the linchpin of electrification, digitalization, and security in the age of AI is both an opportunity and a challenge. The intersection of accelerating demand, constrained supply, and concentrated processing capacity creates systemic risks that require responses from policymakers, regulators, industry, and investors. The choices made in the coming years will determine whether copper remains an enabler of progress or becomes a bottleneck to growth and innovation.

This report seeks to provide a foundation for understanding these dynamics and for charting a path toward the resilient, sustainable copper value chain that is required for the future.

The United States designated copper as a critical mineral in November 2025 reflecting its unique role as the metal of electrification in a world that is being increasingly electrified. The need for copper will grow substantially in the years ahead in order to meet the myriad demands. Yet on its current track, copper supply will fall short with a resulting impact on economic activity, overall growth, and technological advancement. Understanding the complex and challenging interplay of future supply and demand of copper is what this study is about. It is also about understanding how to overcome obstacles in order to facilitate investment and timely development of new supplies.

Copper continues to do what it has primarily done for a century and a half – enabling the lighting of homes and offices and factories, the cooling through air conditioning and the delivery of heat. Today, it is also the connective artery powering physical machinery, digital intelligence, infrastructure, communication, and the movement of people and goods. Ensuring copper’s supply is no longer a question of mining policy; it has become a systemic challenge that touches technology, energy, security, and geopolitics.

In 2022, S&P Global conducted a comprehensive study on The Future of Copper that defined a new era and a new role for copper. For decades, the metal has been dubbed the dour “Dr. Copper” because variations in demand for such uses as construction, lighting, and machinery would be an early warning sign of an impending slowdown or recession. But our 2022 study pointed to a new category of demand – what we called “energy transition demand” resulting from the rise of solar and wind power, expanding sales of electric vehicles, and the drive to electrification as a result of climate policies. In that study, we assessed the substantial additional call on copper resulting from the overlay of energy transition demand on top of core economic demand.

This new study responds to a critical development since 2022 that is driving change – the accelerating pace of electrification. Copper is the essential, highly conductive metal for the endless miles of wiring in buildings and the wiring of every electric motor found in household appliances and industrial machinery, cementing its role as foundational to everyday living. As sectors like construction and manufacturing proceed with digitalization and further electrification, making buildings "smarter" and processes more connected, the demand for copper is intensifying. And while conventional autos use copper, electric cars use 2.9 times more copper, and 22 million of those EVs were sold worldwide in 2025, compared to only about 10,000 just 15 years earlier in 2010.

In this new study, we identify a third category that has only emerged since 2022. This is AI demand, resulting from the surge in investment in “intelligence factories” – otherwise known as data centers – and the infrastructure and ecosystems that support them. The impact is expected to be very substantial. S&P Global estimates that, for instance, the data center share of total electricity demand in the United States will rise from 5% today to as much as 14% by 2030, just half a decade away. This will require a substantial expansion in electric power infrastructure. Indeed, electricity is often described as the major constraint on AI. In turn, the availability of copper could be a constraining factor on expanding that essential flow of electricity (as well as data center equipment itself) – and thus on the roll-out of AI and the promise it holds for efficiency and innovation.

We also point in this new study to yet a further call on copper coming from the global surge in defense spending and the development of new types of weapon systems that depend on advanced electronics, sensors, propulsion, and communication systems. When combined with AI’s defense applications, ensuring reliable copper supply has become as central to national security as it is to industrial policy, AI, and clean-energy transition strategies.

These categories of demand are cumulative. Through 2040, S&P projects global electricity demand to grow nearly 50%, compared to 2025 levels. Every new megawatt and every new line of digital code ultimately depend on copper. AI systems draw vast amounts of power that must flow through copper conductors linking processors, memory, and cooling systems. Electric vehicles require 2.9 times more copper than conventional cars. Wind, solar and transmission systems cannot function without it, and the same is true of advanced weapons and surveillance systems, the effectiveness of which depend on compact, high-conductivity materials.

Strategic criticality

These demands of electrification make copper not just a question of supply but one of strategic criticality. Moreover, copper is essential because the alternatives are limited when it comes to better fit, price, or performance. While aluminum, graphite, or fiber optics can be substituted in some uses, they cannot compete with copper’s superior conductivity, durability, and versatility. Copper consumption can also be reduced through thrifting – using less copper due to design changes or technological advances. Much of the feasible substitutes for copper that can reduce demand for certain end uses are apparently already in use, leaving limited scope for new displacement.

What does all this mean in terms of numbers? The demand section will present our findings on how consumption will drive an unprecedented surge in copper use. This section includes a “teach-in” on varieties of data centers and their impacts on demand.

The supply section assesses the challenges to meet rising copper demand through new mine development, processing, refining, and recycling. Declining ore grades, regulatory and permitting delays, unbounded litigation, and the concentration of smelting capacity – all these have created structural constraints that could lead to persistent shortfalls and sharper competition for supply and possibly supply shocks. The supply section includes a “teach-in” about the processing system, the complexity and challenges of which are not well recognized.

Our study also assesses risks, including the geographic concentration of mining and refining, which drives disproportionate influence over offtake prices and costs while amplifying volatility and exposing vulnerabilities in global supply chains.

This report provides a foundation for policymakers, regulators, investors, industry participants, and the broader public to understand copper’s essential role in the age of AI – as the linchpin of electrification. It sets out how accelerating demand, constrained supply, and concentrated processing capacity intersect. The analysis underscores what can be done to expand mining investment and new supply and to also stimulate innovation and competition in the supply chain.

Copper is among the most critical metals of the 21^st century, essential for a world that is increasingly electric. Across the global economy we see major growth in demand for electricity, whether for power hungry data centers enabling AI, the global shift to electric vehicles, the 2 billion air conditioners that will be installed, or the electrified weapons of the future. To meet the global power demand of 2040, the world will need to build the equivalent of roughly 330 Hoover Dams, or over 650 one-gigawatt nuclear reactors each year between now and then.¹ Copper is the material enabling this massive growth in power demand – unlocking the age of AI and the electrified future of which it is characteristic.

Copper uses are broadening into areas that are both transformative and strategic for the global economy. Our study finds that the call on copper will grow from 28 million metric tons a year in 2025 to 42 million metric tons by 2040 – an increase of 50% above current levels.

Figure 1. Global copper demand by sector (2025–2040)
Million metric tons copper (MMt Cu)

These developments will generate a call on copper that is far higher than the current capacity to deliver the metal without significant adjustments. This signals a fundamental shift in global industrial infrastructure. Later in this study we will turn to how this demand can be met. But, first, why is demand growing?

• Core economic demand – “Dr. Copper”²
• Energy transition and addition
• The explosive growth of AI and data centers
• Defense modernization

Across all four vectors, this growth in electricity relies heavily on copper as its essential conductor. Economic growth; the accelerating pace of electrification; increasing power consumption; the expansion of renewables; and the resulting need to build, modernize, and/or renew transmission and distribution infrastructure – all these will drive major increases in copper consumption in the years ahead. AI, of course, has emerged as the new key driver behind the surge in data center construction, representing growth in copper demand of 2 million metric tons from 2025 to 2040 for IT infrastructure and its associated power generation requirement. And increasing government defense spending focused on electrified equipment – such as drones, telecommunication systems, and advanced missiles – will make copper essential to national security.

Figure 2. Net change in global copper demand by vector (2025 vs. 2040)
Change in demand by sector, MMt Cu

The global landscape for copper demand is rapidly evolving. Between 2025 and 2040, China and the rest of Asia are expected to account for 60% of the world’s copper demand growth, fueled by the swift adoption of electric vehicles, clean power generation, expanded electrification of buildings and infrastructure, and the extension of electrical grids. In North America and Europe, the race to build AI data center capacity and the expansion of solar, wind, and electric vehicles will serve as the primary engines driving copper consumption. Meanwhile, the Middle East is projected to see one of the highest compound annual growth rates in copper demand, approximately 4% per year through 2040, albeit from a smaller base. This worldwide surge will be shaped by policy-driven initiatives, including investments in renewable energy, grid modernization, data centers, and industrial diversification strategies embedded in national development agendas.

On a global scale, the push to renew, expand, and upgrade transmission and distribution lines – as electrification accelerates in countries and regions – will require over $7.5 trillion in grid investments. Copper is central to this transformation, supporting the movement of electricity that drives economies, fuels technological progress, and links infrastructure worldwide. The magnitude of these investments highlights copper’s essential and unmatched contribution to economic progress, energy transition and addition, and digitalization.

Figure 3. Net change in global copper demand by region (2025 vs. 2040)
Change in demand by region, MMt Cu

1. Core economic demand – “Dr. Copper”: Traditional demand sectors which are core to economic growth: construction, cooling, electric appliances, fossil power generation, internal combustion engine (ICE) vehicles,³ machinery, and non-vehicle transportation segments (rail, shipping, aviation)
2. Energy transition and addition demand:⁴ Electric vehicles (EV), battery storage, renewable power capacity (wind, solar), transmission and distribution (T&D)
3. AI & data center demand: Data center ecosystem, including data center operations and power infrastructure to connect data centers to the grid
4. Defense demand: Spending on defense equipment and infrastructure and the roll‑out of new technologies and weapon systems

Later in the study, we point to a likely fifth vector that would emerge in the second half of the next decade – humanoid robots.

This study uses a bottom-up approach to quantify demand at its point of consumption, not production. For example, we quantify copper use in vehicles sold, not produced, to estimate a country or region’s copper demand. An electric vehicle manufactured in China and then exported to a different country would count as copper consumption for that end country, not China. In an interconnected world, finished products containing copper, whether vehicles or appliances, are often traded between countries. Quantifying copper demand from an end-use consumption perspective enables a better estimate of the embedded demand for the metal and the potential shortages or surpluses countries could be facing due to disruptions across the supply chain. As a result, the quantified demand is based on finished demand for copper rather than refined demand or semi-finished⁵ products demand.

Core economic demand (also known as traditional demand) continues to be the backbone of copper consumption in our forecast. Its range of uses is well known and has been familiar for many decades. Falling under this category are the following: 

1. Construction: electrical plumbing and wiring
2. Machinery: motors, generators, and industrial wiring
3. Cooling: heat exchangers and coils for optimal thermal conductivity
4. Electric appliances: wiring and connectors in household items (refrigerators, washing machines, TVs, and lighting)
5. Transport: railways, airplanes, and traditional internal combustion engine vehicles
6. Fossil power generation: tubing and cooling systems, generator winding, and power cables

It is because of the sum economic impact of these varied uses that the metal has earned the sobriquet of “Dr. Copper.” Variations in demand and the price of copper are often seen as signals of economic vitality or recession.

Let’s take Dr. Copper apart, starting with construction. Construction will continue to accelerate globally, expanding copper use. In addition, a combination of rising populations, hot weather, and growing GDPs is driving fast adoption of air conditioning. Rising spending power and broader access to electricity are unlocking new demand for electric appliances, while expansion of industrialization and transportation is further increasing copper demand for machinery and other uses.

Beyond specific segments, core economic demand has historically been linked to broad economic development. Copper is a key material for the generation and transmission of power, which is a direct contributor to economic growth. Access to power enables productivity, supports industrial clusters and transportation systems, and improves living standards. For example, during China’s rapid economic growth in the first two decades of this century, the government’s substantial investments in power infrastructure and urban expansion significantly drove copper consumption to record levels. This illustrates how the demand for reliable power and electrification, facilitated by copper, underpins industrial productivity and economic growth.

Overall, core economic demand globally is forecast to increase by 2% annually, from 18 million metric tons in 2025 to 23 million metric tons by 2040. Construction and machinery continue to be the largest contributors to core economic demand, while demand for ICE vehicles declines due to the growing share of EVs.

Figure 4. Core economic demand for copper (2025–2040)
MMt Cu

Global copper consumption due to construction increased significantly in the 2000s and 2010s, driven largely by the extraordinary Chinese economic expansion that followed its accession to the World Trade Organization (WTO). Twenty million people were moving each year from the countryside to cities; they needed places to live and work, and that turned cities into building sites. Urbanization, rural electrification, and growing car and appliance ownership resulted in burgeoning hunger for the metal. A typical eight-story building uses around 20 metric tons of copper, mostly in wires and pipes. One after another, such buildings were going up across China at a very fast pace. Between 2000 and 2010, China built an average of 3 billion square meters of new gross floor area per year.⁶ To get a sense of the scale and the impact, that is equivalent to adding each year 5 to 6 full New York Cities’ worth of construction⁷ Between 2000 and 2025, China's share of global copper consumption in construction increased from 34% to nearly 40%. Over the same period, India’s share rose from just 4% to 11%. Together, these numbers signal the sheer scale of construction happening in these countries versus other parts of the world.

Building codes around the world are also changing, favoring copper for safety-critical applications. For instance, the International Building Code in the US now emphasizes non-combustible materials for plumbing and heating, ventilation, and air conditioning (HVAC) systems, making copper a preferred choice over plastics. Similarly, the revised British building fire safety standard BS 9991 strengthens requirements for fire-resistant internal components, encouraging copper piping in residential buildings. Electrical codes also continue to mandate copper wiring for its unmatched conductivity and durability, reducing overheating and fire risks. These changes collectively position copper as the safest and most reliable material for modern building compliance. Climate policy is also playing a role in the push for electrification that is driving increased copper consumption. For instance, a pending law in New York State would require all-electric for new homes, banning natural gas for heating and cooking. While controversy surrounds this proposed new law, this may point to a new trend in climate policy that will add to the increase in global electricity demand.

Going forward, Chinese annual building additions are slowing, owing to high vacancy rates and oversupply of housing, but will be offset by increased urbanization and construction in Southeast Asia and India. Copper consumption in construction is forecast to grow by 1.7% annually between 2025 and 2040, from 7.5 million metric tons to 9.7 million metric tons. Asia (including China) accounts for roughly 60% of this total copper in construction consumption by 2040.

Figure 5. Copper demand from construction by region (2010–2040)
MMt Cu

Machinery and other related end uses are a sizeable driver of copper demand historically and into the future. Motors, generators, and associated industrial wiring are used in machines and equipment to produce, process, and transform goods central to the global economy. An example of the continuing electrification of machinery is found in the oil and gas industry, where electric motors are substituting for diesel engines in hydraulic fracturing that produces shale oil and gas.

Altogether, copper’s malleability, durability, resistance to corrosion, and thermal conductivity make the metal a preferred choice for machinery equipment of all kinds.

In addition to industrial uses, this category includes demand for a range of machinery‑based end uses, including non-vehicle transportation systems such as trains, subways, light rail, aircraft, and others, as well as agricultural equipment. Mass transit systems, for example, are a major consumer of machinery-based copper. Modern electric trains, subways, and light rail transit rely heavily on copper for power cables, signaling systems, motors, and overhead wiring due to its superior electrical conductivity and reliability. The transition to electrified public transport further cements copper's role as an essential material supporting the infrastructure of future cities.

Total demand for copper in machinery, including mass transit systems and other industrial sectors, is forecast to grow by 2% annually from 6.8 million metric tons in 2025 to 9.1 million metric tons in 2040.

Copper consumption for cooling equipment and appliances has nearly doubled since 2010 to reach a combined 2.3 million metric tons in 2025. Demand for these end uses typically grows in line with population and economic development. As countries become more populous and average spending power increases, the ability to purchase air conditioning and modern appliances rises. At the same time, the average number of occupants per household drops, leading to a rise in the total number of households and resulting increase in the market size for this type of household equipment. Since 2000, global population has increased by more than 30% to reach 8.2 billion people today, while the average number of people living in each household has dropped from 3.5 to fewer than three. Asia, especially China and India, has been the key driver for rising cooling and appliance demand as populations have grown rapidly and hundreds of millions of people have been lifted out of poverty through rapid economic development.

Lee Kuan Yew, the founder of modern Singapore, once wryly called air conditioning “the single most important invention of the twentieth century” because of what it enables in terms of productivity in the tropics.⁸ Cooling equipment will be a key driver of copper demand, rising 3.4% annually to reach 2.2 million metric tons by 2040. Between 2025 and 2040, the number of air conditioning units globally is expected to rise from 2.5 billion to over 4.5 billion, with continued growth accelerating in the developing world as incomes rise, access to electricity improves, and air conditioning costs come down.

Similarly, demand for appliances including refrigerators, washing machines, televisions, and computers, bolstered by price decreases, will drive further copper demand in the coming years. Copper demand is expected to grow annually by 4.5% for refrigerators, 2.4% for washing machines, and 2.4% for TVs between 2025 and 2040, reaching a total of 1.5 million metric tons.

Figure 6. Copper demand from cooling and appliances (2010–2040)
MMt Cu

One big exception to the growth trends described above is traditional automobiles. In contrast to the other segments of Dr. Copper, copper demand for ICE vehicles will decline in the coming years, falling 5.5% annually between 2025 and 2040. This decline is driven largely by the increasing penetration of EVs into the market and slowing ICE vehicle sales in key regions around the world. As a result, ICE new vehicle sales are expected to drop from 50 million in 2025 to 22 million in 2040. While the EV ambitions of the early 2020s have tempered in Europe and North America, that is not at all the case in the world’s largest auto market, China, and Chinese-built EVs are gaining market share in much of the world.

A typical ICE passenger vehicle contains roughly 25 kg of copper per vehicle,⁹ mostly used in the electrical wiring harness that connects the different features (including HVAC modules, powertrain systems, emission control systems, etc.) and electrical controls. Copper is also used in windings on the alternator and on the low voltage battery.

S&P Global forecasts that the ICE vehicle fleet will peak in 2026. This is not only because of growing competition from electric vehicles, but also because of increasing utilization of vehicles through things like ride hailing and car sharing, as well as longer vehicle lifespans. As a result, copper demand associated with ICE vehicles is expected to fall from 1.3 million metric tons in 2025 to just 0.6 million metric tons in 2040.

The second exception to the general trajectory of Dr. Copper is in conventional electricity generation. The consumption of copper for fossil fuel power generation is forecast to decline as increased renewables lead to fewer fossil fuel-based capacity additions. While global natural gas demand will continue to grow reaching 165,000 billion cubic feet in 2040 given the new rise in power demand, annual capacity additions of all fossil fuel-based power generation peaked in 2010. Going forward, the total fossil fuel-based power additions are expected to decline by -5% per year, from 147 gigawatts per year in 2025 to 66 gigawatts in 2040, due to a combination of policy changes and increased competition from renewable alternatives. A limited increase in new natural gas capacity from 77 gigawatts per year in 2025 to a peak of 90 gigawatts in 2029 is more than offset by a substantial reduction in new coal additions, falling from 68 gigawatts to just 15 gigawatts per year over the same period. From 2030 onwards, annual additions of both sources are forecast to decline through 2040.

While fossil fuel-based power generation covers a range of technologies and configurations, plants typically average roughly 3.3 metric tons of copper per megawatt, primarily used in electrical wiring, generators, transformers, and other conductive equipment. Given the decline in annual capacity additions, copper demand for fossil fuel-based power generation is forecast to fall from 0.5 million metric tons in 2025 to 0.2 million metric tons in 2040 – a 6% average annual decline.

A decade and a half ago, a second major vector of copper demand began to emerge. This was when solar and wind, decades in development, started to become competitive and gain scale, and when electric cars began to appear in showrooms in the United States and China. This emergence tracked the beginning of the electrified future. In the years since, solar panel costs have declined by 90%, largely because of the vast expansion in panel manufacturing in China. Globally, the current manufacturing capacity is now roughly double the global market size. Over the same years, wind turbines have grown in size and capacity. The adoption of these technologies has been advanced by strong government policies, regulations, and subsidies.

S&P Global coined the term ‘Energy Transition Demand’ in its 2022 report, The Future of Copper, to quantify copper demand from the main sectors focused on reducing greenhouse gas emissions directly and indirectly. These sectors included transmission and distribution lines for wider electrification, clean technologies for renewable power generation, and electric vehicle adoption.

Today, these sectors are accelerating not just to address emissions, but to meet the needs of an energy sector pivoting towards a more electrified future. This shift to electrification is aimed at reducing dependence on fossil fuels, with renewables driving new power capacity additions. Global electricity demand is forecast to grow by 2.7% annually from 2025 to 2040 in S&P Global’s Base Case scenario outlook. This shifting energy mix with a greater focus on electricity will drive a substantial increase in future copper demand.

Figure 7. Global final energy demand
Million metric tons of oil equivalent (MMtoe)

Copper’s role in the energy transition arises from its exceptional conductivity (only silver is better as a conductor), durability, and recyclability. It is used in power cables, transformers, inverters, switchgears, busbars, and a range of renewable energy systems including solar photovoltaic (PV) modules and batteries.

Figure 8. Energy transition copper demand by sector (2020–2040)
MMt Cu

S&P Global forecasts that energy transition demand will be the largest source of copper demand growth between 2025 and 2040, requiring an additional 7.1 million metric tons of annual copper demand between now and 2040. This demand will increase from 8.5 million metric tons in 2025 to 15.6 million metric tons in 2040, an annual growth of 4.1%.

Leading the charge in energy transition copper demand growth from 2025 to 2040 are China (+1.9 million metric tons copper), Asia Pacific (+1.6 million metric tons) and Europe (+1.4 million metric tons). These regions will see copper demand growth driven primarily by the wider adoption of electric vehicles and increased spending in renewable capacity additions, driven by regulations, policies, and increased competitiveness of renewable technologies.

Figure 9. Energy transition copper demand by region (2020–2040)
MMt Cu

The world is electrifying. By 2040, we expect that over 21,000 gigawatts of power generating capacity will be operational, producing 48 petawatt-hours (million gigawatt-hours) of electricity. To meet this, the industry will need to add the equivalent of roughly 330 Hoover Dam power plants, 30 of the giant Three Gorges Dam power plants, or over 650 one-gigawatt nuclear reactors each year between now and 2040.¹⁰ Over 92% of the net generating capacity additions in 2024 were renewable – two-thirds of that in one country, China. By 2040, solar PV, wind, and battery storage¹¹ will account for 62% of the installed capacity and 47% of power generation by source. Renewable technologies continue to have lower capacity factors, which force an overbuilding of systems and of capacity. While natural gas and coal will continue to play a significant role in meeting energy demands and maintaining a stable power supply, the increasing integration of renewable energy sources requires a more reliable and upgraded power grid to manage the variability of electricity generated. It should be noted that the recent surge in orders for natural gas turbines and the renewed turn to nuclear power – both fission and now the potential for fusion – could modulate the trajectory of renewables.

Figure 10. Global installed power capacity by technology (2010–2040)
Terawatt (1,000 gigawatts)

Figure 11. Global power generation by technology (2010–2040)
PWh (1,000,000 gigawatt hours)

Copper is used extensively in transmission and distribution infrastructure. While aluminum is often used for overhead cables due to lower cost and weight, copper is preferred for underground applications given its greater conductivity, higher density, and smaller cross section. Copper is also an essential component of grid transformers, with copper windings playing a key role in carrying large currents with minimal energy loss.

Figure 12. Transmission & distribution network system

The share of underground T&D lines is expected to increase as subsurface cables are less vulnerable to weather and fire-related risks, require lower maintenance, and better suit the aesthetic and space constraints of growing urban areas.

While metal intensity will vary with cable voltage, cross-sectional area, and amperage, typical underground transmission lines will use 19,500 kg of copper per kilometer of transmission cable and 3,700 kg of copper per kilometer of distribution cable.

1. The overall growth in power demand and the role of renewables in the power mix create major challenges for the grids, including:
   The increasing age of power grids – average grid ages in key regions range between 20 and 50 years – highlights a pressing need for replacement and modernization
2. The nearly 50% increase in global power demand by 2040 compared to 2022 levels, driven by electrification
3. The necessity to extend and expand the grid to better match generation and consumption centers, and to enhance grid flexibility, which is becoming an increasingly important requirement
4. The presence of connection bottlenecks for renewables – highlighting the need for additional transmission and distribution lines to ease these bottlenecks

To connect growing power generation capacity to consumers, a cumulative $7.5 trillion investment is required in transmission and distribution lines between now and 2040. 

Figure 13. Global T&D investment outlook (2020–2040)
Real 2024 US$, billions

Annually, an average of $130 billion in transmission line investment and $338 billion on distribution line investment will be needed. Roughly 30% of North American and Asian distribution lines are underground, along with 80% of European distribution lines. Copper represents 66% of the total cable weight for underground cables. Overall, the growth in transmission and distribution spending could drive annual copper demand up to 7.1 million metric tons annually by 2040, a twofold increase compared to 2020.

Figure 14. T&D copper demand outlook by region (2020–2040)
MMt Cu

Figure 15. T&D copper demand outlook by segment (2020–2040)
MMt Cu

Figure 16. Copper content in solar PV

Figure 17. Copper content in wind turbines

Figure 18. Global solar PV annual capacity additions 
(2020–2040)
GW

Figure 19. Copper demand from solar PV capacity additions (2020–2040)
MMt Cu

Figure 20. Global wind annual capacity additions (2020–2040)
GW

Figure 21. Copper demand from wind capacity additions (2020–2040)
MMt Cu

Figure 22. Global BESS annual capacity additions (2020–2040)
GW

Figure 23. Copper demand from BESS additions
(2020–2040)
MMt Cu

Figure 24. Global weighted average of copper intensity in passenger vehicles
kg Cu/vehicle

Figure 25. Passenger electric vehicle copper intensity by region (2025)
kg Cu per vehicle

Figure 26. EV share of total light vehicle sales
%

Figure 27. Global annual vehicle sales by powertrain (2020–2040)
Millions of vehicles

Figure 28. Copper demand from vehicle sales by powertrain (2020–2040)
MMt Cu

Figure 29. Primary electricity use per capita (2022)
kWh per capita

Figure 30. Global data center cumulative capacity (2020–2040)
GW of installed capacity

Figure 32. Annual data center capacity additions (2020–2040)
GW capacity additions

Figure 33. Typical data center ecosystem and associated copper intensity

Figure 34. Data center archetype description
mt Cu per MW installed

Figure 36. Data center copper demand by archetype (2020–2040)
MMt Cu

Figure 39. Global defense spending (2010–2040)
Real 2024 US$, trillions

Figure 40. Defense industry copper demand (2020–2040)
MMt Cu

Figure 41. Copper substitution possibilities

Figure 42. Total copper market balance (2020–2040)
MMt Cu

Figure 43. Three areas of focus for the copper supply chain challenge

Figure 44. Copper supply chain overview

Figure 45. Copper mined production from operating assets (2010–2040)
MMt Cu

Figure 46. Copper production (2025)
MMt Cu

Figure 47. Copper reserves & resources (2025)
MMt Cu

Figure 48. Copper reserves & resources (2025)
MMt Cu

Figure 49. Baseline copper supply and demand outlook (2010–2040)
MMt Cu

Figure 50. Incremental 2040 production by country
Thousand metric tons Cu

Figure 51. Copper average head grade for major copper mines (2000–2024)
% Cu head grade

Figure 52. Comparison of South America average mine site cost vs. head grade (2000–2024)
Left axis: US$ ¢/lb, right axis: % head grade

Figure 53. Selected copper deposit discoveries (1985–2025)
MMt Cu

Figure 54. Copper exploration budget by project state (1997–2024)
Real 2025 $US, billions

Figure 55. R&R change between 2005 and 2023, including produced reserves
MMt Cu

Figure 56. Mine cost curve over time from costed operating mines (2010 vs. 2019 vs. 2025)
Total cash cost ¢/lb

Figure 57. Average depth of drilling activity (2010–2025)
Depth (meters)

Figure 59. Overview of copper recycling

Figure 60. Circular economy in the copper supply chain

Figure 61. Global copper stocks in use (1960–2020)
MMt Cu

Figure 62. Global end-of-life scrap availability from 2020–2040
MMt Cu

Figure 63. End-of-life recycled copper scrap supply by sector (2020–2040)
MMt Cu

Figure 64. Prices of refined copper and copper scrap (2004–2024)
Real 2025 US$ per Mt Cu

Smelting and refining constitute a single point of strategic criticality for the copper market, linking global supply with demand. The high regional concentration of facilities makes the overall supply chain sensitive to volatility or disruption, despite existing overcapacity. During mineral processing, the smelting process is used to extract copper from copper concentrate (typically containing 20-30% of copper) by heating and melting, which separates the metal from impurities. This step is followed by refining, which further purifies the copper typically through electrochemical methods.

These processing steps sit between the mined ore and the final product, making them vital nodes in the supply chain. While mined copper production has grown steadily, smelting and refining capacity has changed unevenly. The ability for mined supply to meet end-use demand increasingly hinges on whether the smelting and refining industry can overcome challenging economics and increase investment through 2040, as a processing deficit looms.

Figure 66. Copper market supply chain breakdown (2025)
Illustrative

While both the US and China will be major consumers of copper in the years to come, the two countries’ trajectories on supply and processing are quite different. While limited supply (mostly through recycling) and processing will cause the US to continue to rely heavily on imports of final copper cathode material, China’s scaling of processing capacity provides it with more sourcing options from concentrate suppliers.

Figure 67. Copper supply chain comparison, US vs. China (2025)
MMt Cu

In the last 10 years, the growth in smelter capacity has outpaced new mined concentrate production. The industry has seen an increase of 8 million metric tons of new smelting capacity, whereas copper concentrate production has increased by only 3 million metric tons in the same period. These dynamics have created increased competition between smelters for access to concentrate, putting pressure on treatment charges. The result is to crowd out less competitive smelters.

Figure 69. Smelting copper capacity by region (2009–2028)
MMt Cu

It is estimated that to build a smelter in the West would cost somewhere between one and four billion dollars, depending on the facility. Moreover, in many jurisdictions, a new smelter would likely face considerable challenges in the permitting process and in courts. Changes in the copper smelting environment have been largely driven by increases in Chinese smelter capacity from 2000 onwards. Chinese smelters now represent roughly 40% of total smelting capacity and dominate the imports of the main input, mined copper concentrate. This makes them increasingly important in annual treatment charge negotiations with copper miners. In the same period, the market share of the top three producing countries for mined copper – Chile, DRC, and Peru – has decreased.

Figure 70. Regional concentration (2018–2028)
% of total

It is estimated that to build a smelter in the West would cost somewhere between one and four billion dollars, depending on the facility. Moreover, in many jurisdictions, a new smelter would likely face considerable challenges in the permitting process and in courts. Changes in the copper smelting environment have been largely driven by increases in Chinese smelter capacity from 2000 onwards. Chinese smelters now represent roughly 40% of total smelting capacity and dominate the imports of the main input, mined copper concentrate. This makes them increasingly important in annual treatment charge negotiations with copper miners. In the same period, the market share of the top three producing countries for mined copper – Chile, DRC, and Peru – has decreased.

Figure 71. Illustrative smelting revenue structure

In a rush to secure feedstock and capture those other revenue sources, TCRCs have dipped below zero. In these unusual circumstances, the smelter pays the miner to process the ore and keeps the by-product revenues from other metals contained in the ore and sulfuric acid. Although some smelters can withstand the cost in the short term, difficult economics could force smelter closures.

Figure 72. Spot TCRCs (2021–2025)
¢/lb (CIF Shanghai)

In this market, smelters have a lower ability to absorb the shock of decreasing revenues as they do not benefit as much from direct price exposure to copper as miners would. If smelting capacity contracts just as demand for refined copper accelerates, the constraint shifts from mine to furnace, delaying metal availability even if the concentrate is produced.

Figure 73. EBIT margin of copper smelting and mining companies (2010–2024)
%

The economics of smelters, while currently pressured by low treatment charges, vary significantly across regions. Operating costs are influenced primarily by electricity, fuel, labor, and maintenance/utilization rates. China-based smelters tend to have structurally lower operating costs than other regions, driven by relatively competitive industrial electricity rates, access to skilled lower-cost labor, and modern capacity with less need for maintenance. These lower costs help make China-based smelters more resilient in a volatile market.

Moreover, other factors are essential in shaping competitiveness. First, the ability to negotiate TCRCs is key. In China, the role of the China Smelters Purchasing Group (CSPG), which coordinates negotiation positions among major smelters, can amplify bargaining power during annual benchmark discussions.²⁶ This collective approach helps large Chinese smelters secure more predictable terms and reduce the volatility. Second, clustering effects further strengthen competitiveness: the close concentration of smelting, refining, semi fabrication, and final good manufacturing facilitates logistics costs and talent sharing. As the copper market and supply chain continue to evolve, these factors will play a role in how the smelting and refining players adapt to meet growing needs.

Finally, technology matters. Modern copper smelting technologies, particularly those utilizing flash smelting or IsaSmelt²⁷ processes, deliver notable improvements in efficiency and process control compared to reverberatory furnaces still operating in select regions. These advanced methods enable production of cleaner copper anodes and facilitate enhanced recovery of by-products – including gold, silver, tellurium, and selenium – from copper anode slime during subsequent electrolytic refining. Additionally, newer smelters are engineered for greater thermal and chemical adaptability, supporting the economical processing of a broader spectrum of feedstocks. This includes not only lower-grade copper concentrates but also mixed, complex, and lower-value copper scrap. As a result, these facilities are better positioned to access secondary material streams, whereas older, less flexible US smelters – primarily designed for high-grade ore – face limitations in processing such inputs efficiently. Large capital investment would be required to process lower-grade scrap.

In itself, copper is roughly a $250 billion business. But of course, it is much more when one considers that so much of the goods that are shipped around the world depends on the “copper inside”.

The supply chains that carry copper span the globe, as discussed in the chapter above. They link the long, uncertain process of exploration and the complex scale of mining with the industrial systems and advanced technologies that depend on copper at every stage of value creation. Figure 74 traces the flow of copper from primary production through the supply chain and the role of trade across each step.

Figure 74. Trade at major stages of the global copper supply chain

Risk accumulates where the flow is most concentrated. On the export side, a limited number of countries dominate concentrate production and shipments. On the import side, ore and concentrate imports are even more concentrated, with a small group of large processors purchasing a disproportionately high share of globally traded feedstock. This dual concentration has significant impacts: it shapes pricing power, determines the structure of offtake contracts, and locates where value is captured along the chain. In 2024, global mine output of concentrate lagged available smelting capacity by roughly 1.5 million metric tons of copper content, creating a global competition for copper concentrate imports that pushed smelter TCRCs to multi-year lows. The consequence was to shift bargaining leverage toward miners. Because new copper mines average 17 years from discovery to full production, this imbalance between mine output and processing demand is likely to persist.

Modern smelters are designed to operate at high utilization rates even when TCRCs are compressed. Several structural advantages support this: competitive industrial power pricing; scale efficiencies in large integrated complexes; access to low-cost finance; and – of importance – the ability to process a broader range of mineral-bearing concentrates.

This last capability is increasingly important. Many large smelters can process “polymetallic” concentrates – that is, concentrates that contain other minerals – gold, silver, molybdenum, and minor metals such as selenium and tellurium. These by‑product metals provide additional revenue streams that allow processors to offset low copper TCRCs with by-product credits. Figure 75 illustrates how investment in modern smelting capacity, combined with lower energy costs and economies of scale, have concentrated refining capacity in dominant processing hubs. Between 2015 and 2024, China’s share of concentrate imports rose substantially from 43% to 66%, in line with China’s expansion of smelting capacity mentioned in the chapter above. The share of all other major importers declined. These polymetallic smelters, operating at scale, have a decisive competitive advantage: they remain profitable at TCRC levels that make new or higher-cost smelters elsewhere difficult to finance or operate.

Figure 75. Largest global copper ore and concentrates importers
Percentage share of imports (%)

When concentrated, modern smelting capacity commands the largest share of ore and concentrate imports, commercial decisions exert outsized influence on global benchmarks and contract structures. Persistent low TCRCs discourage new smelter projects in regions with higher operating costs or stricter permitting, reinforcing geographic concentration. This raises barriers to entry, increases systemic exposure to disruptions in a few locations, and limits the development of more diversified processing capacity.

These dynamics extend upstream and downstream. Manufacturers of transformers, motors, cables, and data-center components depend on predictable access to refined copper and semi-finished goods. Upstream, miners make exploration and development decisions based on long-term price expectations and offtake terms. If control of copper trade accrues at the processing stage – through tight control over concentrates and by-product capture – such concentration can weaken exploration spending, slow the pipeline of new deposits, and perpetuate the imbalance between mine output and processing demand.

The turmoil in the rare-earth sector offers a structural parallel. Over time, separation and processing capacity became highly concentrated. When export controls as part of larger trade tensions were later applied to rare-earth oxides and magnet materials, the impact fell directly on industrial processes in importing countries, which lacked alternative routes for processing. A wide range of manufacturing capabilities were abruptly put at risk. Short-term price movements had masked a deeper vulnerability. Copper has important and different characteristics from rare earths, but there is an underlying lesson: when mine development is slow and processing is concentrated, policy shocks at the processing stage can propagate quickly across global manufacturing chains.

Tariffs are often introduced to encourage domestic production, stimulate investment, or protect strategic industries. In copper, however, their effects depend heavily on where they are applied along the supply chain and how they interact with geology, project timelines, cost structures, and existing industrial capacity.

There is nothing simple about tariffs. Managing tariff policy is complicated by the structure of the tariff system itself. Tariffs are defined at detailed product levels –for example HS-6 (the internationally harmonized six-digit product classification) and HS-8 (national extensions to greater specificity). They are then modified by free‑trade agreements, anti-dumping and countervailing duties, safeguard measures, and national-security actions. Exemptions, quotas, and temporary executive orders further alter effective rates. The result is a highly complex structure in which the actual tariff applied depends on precise product classification, origin, and the overlay of special measures.

High tariffs on imported ores or concentrates raise costs for domestic processors and manufacturers without generating new mine supply in the near term, given the long timelines for new projects. High tariffs on semi-finished or finished copper products, even with tariff-free ore imports, can inadvertently reinforce the processing and manufacturing advantages of competing regions with lower operating costs and established smelting capacity.

Market shocks in July 2025, based simply on rumors of impending US tariffs on copper, illustrate the sensitivity of market perceptions of tariff structure. When the US Administration announced plans for a 50% tariff on copper imports, markets initially interpreted the measure as applying to raw refined copper. That interpretation triggered an immediate surge in prices in the two major copper exchanges. In the COMEX (Commodity Exchange Inc.), the US futures exchange where copper is traded primarily as a financial instrument, copper futures spiked sharply. In the LME (London Metal Exchange), the home of the global benchmark exchange for physically deliverable industrial metals, prices rose, and the spread with COMEX widened significantly – signaling differences in regional supply-demand conditions, financing costs, and logistics, and a tighter US market.

Once an official announcement made clear in July that the 50% tariff would apply primarily to semi-finished copper products and copper-intensive derivatives, not to copper concentrate, prices corrected and spreads narrowed. A dramatic point was made: expectations alone – even before final tariff schedules are published – can distort price discovery and affect hedging, contracting, and supply-chain planning.

Figure 76. Increase of US input costs and output prices for selected electrical components
% CAGR, 2020Q1 to 2025Q2

In the same vein, tariffs on semi-finished copper can amplify downstream pressures and distortions. Figure 76 shows that from 2020 to 2025, US transformer prices rose nearly 45%, reflecting raw-material inflation and the July 9 tariff pass-through. Similar increases occurred in switchgear, motors, generators, and panelboards – essential equipment for grid modernization, electrification, and digital infrastructure. When tariffs are not aligned with the supply-chain competitiveness or the capacity to respond, they can raise domestic product costs without necessarily diversifying supply.

A world in which tariffs play a more significant role will have many and varied impacts. Depending on where they fall, they can encourage or discourage investment. They can provide greater predictability or they can increase uncertainty. They can help producers and impose costs on consumers or vice versa. They can benefit some segments in the value chain and disadvantage others. They can be put on or taken off. They can certainly change how copper in all its forms – from rocks to final products – flows around the world. They can make themselves felt in direct tariffs related to copper or on the products that embody copper that are shipped across the globe. One thing does seem certain. If the current trend continues, tariffs will be among the significant risks and uncertainties for the global trade in copper.

While the need to increase global production of copper is ever clearer, several above‑ground challenges persist. This chapter examines these challenges, which shape the development of copper mines, and the lessons that emerge from them. A central challenge is time: as detailed below, the long and lengthening timelines required to discover, assess, permit, finance, and construct a modern copper mine have become one of the most significant constraints on project viability. Extended development cycles raise capital requirements, delay cash flow, and increase exposure to regulatory or political shifts. In the context of the emerging shortfall in global copper supply, these risks are not just commercial – they directly affect the world’s ability to secure the copper needed for electrification, grid modernization, digital infrastructure, and the broad productivity gains that accompany copper-intensive applications.

On average, a new copper mine takes 17 years from first discovery to first production. More than two-thirds of that time – over 12 years – is consumed by early exploration, feasibility studies, environmental assessments, and legal or administrative delays or challenges in obtaining permits.

Figure 77. Development times for copper mines, global average

These timelines are getting longer. Projects that began production in the late 2000s took on average 13 years to develop, while those that began production in the early 2020s typically took 18 years.²⁸ Longer development times are widely attributed to new and expanded environmental challenges and permitting requirements, especially in countries that require baseline feasibility studies of impact on communities and the environment. The landscape of consultations with local communities is expanding, addressing concerns that span highway safety, water scarcity, waste management, and employment opportunities. Navigating the intricate web of national, sub-national, and local laws presents its own challenges.

Furthermore, the involvement of a wide range of agencies and regulatory bodies, whose authorities frequently overlap, adds complexity. At each stage of the permitting process, there remains the potential for legal intervention by opponents, underscoring the multifaceted nature of these engagements. All these add to the scale and complexity of permitting (see the “Permitting in the United States” case study below). Defined processes (environmental, community engagement, etc.) are important for the sector, but standardization is key as differences across jurisdictions create delays and uncertainty for investors.

Across Latin America, for example, community consultation has become central to mine development, but the boundaries can be unpredictable. National frameworks – many influenced by International Labour Organization (ILO) Convention 169 – vary widely: Mexico and Chile use non‑binding citizen consultations, while Peru’s binding Indigenous consultations can halt or reshape a project. Experience has shown that well-designed consultations can strengthen legitimacy, long-term stability, and community partnership, including job creation. But when the scope, sequencing, or duration of consultations is unclear, uncertainty grows – and multi-year delays can undermine project viability and eventually constrain copper supply. There is a further complication: intervenors may have different motivations. Some may be concerned about road congestion and water availability. Others may be opposed to mining for environmental, ideological, or political reasons. Some may be seeking a source of revenue. Thus, there is a practical challenge: how to uphold meaningful consultations while providing predictable, time‑bound processes that give all stakeholders clarity.

The scale of modern copper mines compounds these challenges. Copper deposits tend to be lower grade and more diffuse than most other metals (see Chapter 3.1 Primary supply). Larger sites and more extensive drilling are required to define viable deposits. Bulk-tonnage mining – extracting vast quantities of material to achieve economies of scale – may necessitate significant investment in infrastructure such as new access roads (especially for remote sites), power connections, water systems, and waste management. All these demand comprehensive planning and assessment, triggering additional reviews. 

The copper industry is grappling with a structural shift to a higher cost base, driven by a confluence of persistent inflation, deteriorating geological fundamentals and escalating capital requirements for new projects. It takes more equipment, mining to farther depths, to extract and then process the volume of ore needed for commercial viability. For example, the global ratio of waste rock to ore in copper mines, the ‘stripping ratio,’ has increased from 1.47:1 in 2021 to an estimated 1.80:1 in 2025. This raises both the capital and operational expense of copper mines.

In recent years, general inflation has further increased operating costs by driving up the prices of equipment, labor, and energy. Costs have risen for diesel, natural gas, and electricity needed for operations. From 2021 to 2024, the total cash cost per ton of ore treated surged by a cumulative 28%, from $23.6 to $30.2 per metric ton. This was mirrored by all-in sustaining costs (AISC) on a co-product basis, which climbed 23% to 268.7 cents per pound. While costs will stabilize as global inflation does, they are likely to do so at this permanently higher level.

The copper industry needs to meet future demand for the metal, estimated at 42 million metric tons by 2040, as identified in Chapter 2 Copper demand in the age of AI above.

Building new copper mines is expensive and slow because of many physical, regulatory, and bureaucratic challenges. S&P Global looked at the capital and operational costs of new and expanding mines to understand what copper price would be needed to encourage more development.³¹

Figure 79 shows that, even if all the listed mining projects overcome the necessary barriers to proceed and do not face more cost increases, the 2025 average copper price of $9,500 per metric ton is only high enough to make roughly 60% of projects profitable, leaving the remaining 40% potentially uneconomical.

Figure 79. New copper projects’ capacity ranked on total incentive price, 2035
Total cost, thousand US$/ metric ton Cu

The growing capital intensity and cost of new copper projects carries implications that extend far beyond the mining sector. Higher development costs translate into elevated copper prices, and those prices ripple through the economy in ways that are often underestimated. Electrical infrastructure becomes more expensive to build, renewable energy projects face rising installation costs, and data centers see an increase in capital expenditure per megawatt. For example, a 2,500 kVA power transformer can contain over one metric ton of copper – about 30% of its total mass – making it highly sensitive to copper price movements. What begins as a challenge of project economics evolves into a constraint that affects the affordability of electrification, the pace of digitalization, and the economics of energy transition. Copper is not only a physical bottleneck; it is becoming a financial variable that shapes investment decisions and influences how quickly critical projects move forward. Copper prices moved into a higher range during COVID-19 and have recently surpassed $11,500 per metric ton.

Figure 80. Spot copper LME price
US$/metric ton Cu

Copper's role in the new era of AI data centers highlights its strategic criticality, as the metal alone accounts for a significant, multi-million-dollar component of the total capital expenditure of a single data center. For a large, greenfield AI training data center – like one with a 230 MW capacity and an estimated total cost of $3 billion—the massive quantity of copper required for power, cooling, and connectivity systems translates directly into a substantial cost. Based on its high-intensity use of 44 metric tons of copper per MW and the December 2025 spot price of $11,500 per metric ton, the sheer volume of copper needed for such a facility runs to nearly 10,000 metric tons, pushing the total material cost of copper to nearly $115 million. 

Mining is inherently a long-term, capital-intensive investment, often requiring billions of dollars and decades of commitment before reaching full production. Because public budgets rarely have the capacity to finance projects of this scale nor public authorities the capabilities, governments generally depend on private capital to develop mines and supporting infrastructure. To secure such investment, governments at the beginning typically offer stable tax regimes, competitive royalty frameworks, predictable permitting procedures, and balanced local-content obligations – assurances meant to give investors confidence in a project’s long-term viability. Yet once capital is sunk and construction is well underway or the mine operating for a couple of years, bargaining leverage shifts: the investor becomes committed, the host government less constrained, and the incentive to revisit earlier commitments increases. This shift in bargaining power – after major investment is irreversible – is the essence of the ‘obsolescing bargain’.³²

Figure 81. Government action and resolutions of recent mine projects 

Across Africa, Latin America, and parts of Southeast Asia, successive renegotiations have increased perceived risk, raising financing costs and delaying or cancelling investment. While some countries maintain consistent legal and regulatory frameworks, others experience abrupt policy shifts when governments change, or political budgetary pressures intensify, or new priorities appear. These shifts often take the form of updated royalty regimes, new export restrictions, expanded local-content mandates, or contract revisions justified as efforts to enhance national benefit or assert greater economic sovereignty. Such unpredictability reinforces investor concerns and slows the development of new copper supply.

The core challenge is that mining and politics run on different timelines. Mines unfold over decades, while many governments operate on multi-year election cycles that constantly reshape incentives. Over time, governments and mining operators have learned that bridging this gap depends on continuous engagement, clear understanding of parallel co-benefits that projects may deliver to communities, and governance arrangements resilient enough to endure political shifts – all essential to maintaining stable investment conditions in a sector critical to global electrification and growth.

Another emerging risk can inhibit development of new supplies: lack of skills and experience going forward. From exploration to mining to processing to permitting, mining demands advanced technical expertise. Geologists must locate and characterize deposits. Geophysical surveys and chemical sampling are used to delineate resources, which must then be accurately modelled. Mining engineers are essential to design, plan and operate mines safely, with close consideration of ground support, slope stability, ventilation, haulage routes, blasting sequences – all to exacting safety and environmental standards. Tailings and waste engineering must take account of a range of environmental regulations, local concerns, and operational efficiency.

The talent gap takes the form of inadequate numbers of appropriately skilled young people coming into the industry and government. Mining companies depend on these skills at every stage of exploration, planning, permitting and operation. Staff in government permitting agencies require similar expertise and experience to secure and sustain investments and fulfill their public obligations. Access to these skills is declining globally.

Retiring career experts are not being replaced by new graduates. The Centre for Strategic & International Studies estimated that 221,000 US mining workers will retire by 2029.³³ While the vast majority of these will not be professional engineers, this figure compares with just 327 degrees awarded in 2020 in mining and mineral engineering programs. The number of programs themselves is falling. There were 25 such programs offered in the US in the early 1980s, falling to 15 by 2023. (In contrast, Chinese institutions offer 38 mineral processing schools and at least 44 mining engineering programs.)

Looking to the future of mining, people skills are indispensable to deliver the innovation that can boost metal production and integrate technological advances into the exploration, engineering, and production phases. Building on current opportunities and integrating new advances depend upon a reliable supply of trained geologists and engineers.

While policy has some traction on the complexities of permitting, the stability of fiscal regimes, and even the development of talent, other above ground challenges are difficult to control. Deepening water scarcity is emerging as a significant limitation on a water-intensive industry (dust suppression in open-pit mines; crushing and grinding ore; separating copper minerals from rock; and cooling, among other processes). In the largest copper mining country, Chile, the industry’s need for water is increasingly set against that of agriculture and households. In 2023, the Cerro Colorado mine’s license to extract water from the Laguillas aquifer was not renewed in light of other demands on the aquifer. Solutions are available – but costly. The Los Pelambres mine built a large desalination plant in 2024 to reduce its dependence on dwindling freshwater sources. While the plant has so far significantly raised the project’s ore processing rates, it required substantial new capital to be sunk. In the medium-term, the retreat of glaciers in the region worsens water scarcity and introduces further geological challenges including permafrost melting, landslides, and rockfalls.

The impact of all these above ground challenges is making mining more costly, raising risks and making operations more politically fraught. That raises the bar for the capital required to develop and sustain productive, safe, and efficient mines with social license to operate. Investors must be prepared to sink huge sums of capital, engage a multitude of stakeholders, wait sometimes for decades for returns, and ringfence further capital for the various contingencies – from legal challenges to rockfalls – that mines face. Without this investment, however, the world will not be supplied with the copper it needs. 

A growing number of governments recognize that in the age of critical minerals, stability and competitiveness can matter as much as geology. The approaches taken vary widely, but they all reflect an effort to replace unpredictable fiscal extraction with clear, long-term partnership frameworks.

Argentina has taken one of the boldest steps in this direction. The Régimen de Incentivo para Grandes Inversiones (RIGI), mentioned above, was enacted in July 2024 and implemented by decree the following month. It provides a comprehensive framework for large-scale projects in strategic sectors such as mining and energy. Within just six months – by January 2025 – the government approved its first project under the regime, the Los Azules copper development, valued at roughly $2.7 billion and expected to generate more than $1 billion in annual exports. RIGI consolidates exploration, construction, and operation under a single legal and fiscal framework, offers 30-year tax and customs stability, reduces corporate income tax from 35% to 25%, exempts VAT during construction, and eliminates export duties after the third year of operations (or the second for projects above $1 billion). The speed from legislation to the first approval demonstrates the government’s intent to compete aggressively for global investment through predictability and incentives rather than protectionism.

A second model can be seen in Japan-Australia cooperation on critical minerals, which has become a benchmark for allied industrial coordination. Through public-finance institutions such as the Japan Bank for International Cooperation (JBIC) and Japan Organization for Metals and Energy Security (JOGMEC), Japan has partnered with Australia’s government and firms like Lynas Rare Earths to secure stable supplies of rare earths, lithium, and copper. The partnership combines long-term offtake agreements, concessional finance, zero-tariff trade frameworks, and shared environmental and transparency standards. In 2024–2025, both countries expanded this collaboration under the Japan-Australia Critical Minerals Partnership, establishing joint processing hubs in Western Australia.

In some cases, resource-holding countries are pushing for downstream investments in processing. But, as demonstrated above, smelting and refining are not necessarily profitable businesses. They can create jobs, but their risks may well be an economic net loss to the country and lead to concentration of ownership of these downstream assets.

For all countries, it is important to consider the prospect of legal risks. The number of stages at which cases can be brought affects cost, uncertainty, and delay. Reducing the scope for legal challenges while protecting recourse to courts for legitimate complaints could significantly reduce mine development times.

Copper has transcended its traditional role as an economic prognosticator. It is no longer just "Dr. Copper", the metal whose price would signal the ups-and-downs in the economy. It is also the metal of this new era of electrification.

The world is entering a time in which economic demand, grid expansion, renewable power generation, AI computation, digital industries, electric vehicles and defense – all these – are scaling all at once. As a consequence, global copper demand is on course to surge 50% by 2040, not only along the vector of the familiar markets that have built the modern world, but also along these new vectors of energy transition, artificial intelligence, data centers, and defense modernization and battlefield electrification. And there is a possible new vector that could become clearer by the beginning of the next decade – the arrival of humanoid robots.

Power consumption is accelerating across every major region, with global load on track to grow by half by 2040. The connective ligament of the expanding power system is copper.

The challenge ahead is that the accelerating pace of electrification now exceeds the pace at which copper supply is set to grow. As a result, the unprecedented increase in demand for electricity confronts a sobering reality: a potential 10 million metric ton supply shortfall — 25% below projected demand — that threatens to constrain global technological advancement.

Primary production – mining – remains the irreplaceable foundation. Bridging this gap demands an extraordinary, multi-dimensional response. Future copper supply depends not only on geology, engineering, logistics, and investment, but also on governance and policies. That translates into timeliness in permitting and consultation, a time clock on litigation, and stability in governance and regulation.

All of this is required to engender the confidence to underpin the substantial investment that spans decades. Long and jagged timelines magnify the cost of uncertainty. Host countries that reach clear agreements in a timely fashion, define terms predictably, and coordinate decision-making processes will attract investment and bring supply forward. Uncertainty comes with a cost. Countries that underplay transparency and predictability will face slower development, weaker community benefits, impacted revenues, and higher risk premiums on capital.

Yet mining itself is only part of the picture. It is also about what happens to the copper concentrate when it leaves the mine. Ensuring robust supply chains and diversified processing capabilities have become central priorities. Processing – smelting and refining – is a critical node in the supply chain, especially as capacity is concentrated in a limited number of countries. Building a more resilient global copper system requires multilateral cooperation and more regional diversification. A wider base strengthens resilience, fosters better environmental performance, and ensures that both advanced and developing economies share in the benefits of the electrified era.

The future is not just copper-intensive, it is copper-enabled. Every new building, every line of digital code, every renewable megawatt, every new car, every advanced weapon system depends on the metal. As electrification and digital intelligence become defining characteristics of global development, copper is indeed an ever-more critical mineral, carrying the electric currents that are connecting, conducting, and catalyzing innovation and economic advance.

Data centers are specially designed facilities that concentrate the computing, storage, and network infrastructure that powers the digital economy. Data center buildings house servers, along with power, cooling, and security systems. They provide core services ranging from cloud and enterprise information technology (IT) to applications in AI. Given their role in today’s world, data centers are critical to everyday life and must deliver a very high level of reliability to their users.

While data centers have been around since before the turn of the millennium to serve cloud services, interest in data centers and their associated power consumption has exploded after the release of widely accessible AI tools to the public. For AI use specifically, data centers rely on power hungry GPUs. With each new generation of GPU, power demand increases. As an example, in our modeling, data centers globally could consume more power in 2030 than what Latin America consumes today. Consequently, the main constraint for data centers is power availability, and thus the resulting copper intensity.

Data center development varies widely across regions. Key to enabling development are factors including power availability and reliability, land, access to markets, and the supply chain. Given that power is both a major cost and primary constraint to data centers, locations with reliable and readily available power are the main enabler for data center growth. While having access to markets through fiber cables is essential for some data centers, others need to be in close proximity to markets due to latency. Other key enablers include access to water (for cooling), favorable regulation, clean power and land availability.

• Computing: Servers that process data and execute workloads
• Storage: Systems that retain data
• Networking: Infrastructure that connects systems internally and externally to deliver data to end-users

Data centers come in several forms, each designed to meet specific operational needs, redundancy requirements (for critical use-cases), computing needs, and tailored services. The main data center types include:

• Hyperscale: Facilities built by tech giants (such as Alibaba, Alphabet (Google), Amazon, Apple, Baidu, Meta, Microsoft, Tencent), typically deployed as multiple facilities on a campus.
  Capacity size: 50-100+ MW.
  Key characteristics: Designed for extreme scalability and efficiency, can handle regular cloud computing and high-density chips for AI workloads.
• Leased: Third-party facilities built to be leased to one or multiple tenants, typically using one of two business models – retail or wholesale.
  Capacity size: 10-50 MW.
  Key characteristics: Reduces capital expenditure for customers, who do not own the infrastructure. Top leased data center customers are hyperscale IT firms who build their own data centers but also lease facilities to improve time-to-market.
• Enterprise: Self-built facilities by enterprises, universities, or governments that tend to be deployed near corporate offices or customers.
  Capacity size: 1-10 MW.
  Key characteristics: Often highly redundant and tailored to specific business requirements.
• Crypto: Facilities used for crypto mining and flexible regarding location, as they typically are deployed where power is less expensive.
  Capacity size: 1-10 MW.
  Key characteristics: Used for crypto mining and blockchain transaction validation, some “host” facilities house machines used by other crypto miners.

The workloads hosted in a data center determine its design, as they can have different requirements in terms of latency, redundancy, hardware, and compute power. Latency in this context refers to the network delay between a request and the corresponding response and is minimized by locating data centers close to end-users. A lower latency means faster response times, which is critical for trading, streaming, and autonomous systems, while applications with higher latency can tolerate delays in data transmission.

Chips are the main hardware used in the servers located in data centers and are categorized by their computing capabilities. GPUs are specialized, high-performance chips that are used in AI modeling, as well as gaming, and consume significantly more energy than traditional general-purpose chips such as Central Processing Units (CPU).

The majority of data center power goes to IT equipment, which includes servers, data storage, and networking. The second-largest share goes to cooling that equipment, with a small percentage (5-6%) coming from power distribution losses and lighting. Power Usage Effectiveness (PUE) is the standard metric used to assess the power efficiency of a data center. It is the ratio of the total amount of power needed to operate the data center, including cooling and lights, divided by the power required just to run the IT equipment. The lower the number, the less additional power is used to run the facility beyond what is required for servers and networking, and the more efficient it is. Traditional enterprise facilities tend to have PUEs around 1.8 to 2.0 due to older hardware, inefficient cooling and low utilization, while hyperscale data centers can approach a PUE of 1.1 to 1.3.

Data center AI workloads can be mainly categorized as training or use/inferencing. AI training involves large-scale computing that feeds data into algorithms to help an AI model iteratively learn patterns and relationships. Training requires machine learning algorithms and significant computing power but tends to not be latency sensitive, as the data processing is relatively self-contained. This means the data center can be in more remote locations to alleviate pressure on capacity-constrained electrical grids. AI inferencing/use requires less computing power than AI training and is often not latency-constrained and therefore can, in some cases, be in remote locations. However, there are some AI inferencing use cases that require low latency or interaction with a large amount of data that is not part of an AI training facility. In this case, the AI inferencing center may need to be located close to end-users or to the data required.

AI training data centers are often purpose-built to handle large-scale models. In contrast, non-training data centers have the flexibility to support what are now traditional cloud services, web hosting or enterprise applications while also running AI inference, balancing flexibility, efficiency, and latency requirements for a broader range of services.

Figure 82. Data center details

Figure 83. Copper demand
MMt Cu

Figure 84. Baseline copper supply
MMt Cu

Figure 85. Unconstrained copper supply
MMt Cu

The data in Figure 85 represents unconstrained supply from committed, probable, possible, and speculative projects. Speculative projects comprise assumed potential volumes from identified sources with a high degree of uncertainty (up to 5 MMt by 2040), intended to account for potential expansions, new discoveries, and other developments not included in S&P Global’s project list. To estimate the baseline (constrained/risked) production, disruption rates of 90% to 100%, 65%, and 40% are applied to forecast mine output from committed, probable, and possible projects, respectively.

Figure 86. Acronyms and their definitions