> **来源：[研报客](https://pc.yanbaoke.cn)** # From Mineralsto Megawatts: # Building Resilience for EVs, Data Centres and Power Grids WHITE PAPER DECEMBER 2025 # Contents Foreword 3 Executive summary 4 1 Understanding mineral intensity 5 1.1 The known frontier of electric mobility 7 1.2 The hidden footprint of digitalization 9 1.3 Copper and steel for electrification 11 2 Supply chains disjointed by tiers, territories and timelines 13 2.1 Anatomy of the electric mobility supply chain 14 2.2 Structure of the digital infrastructure supply chain 16 2.3 Map of the grid supply chain 18 3 Shared risks, divided attention 20 3.1 Early warning signs in supply 20 3.2 Upstream shocks, downstream consequences 24 3.3 Competing risk agendas 25 4 From awareness to action 29 4.1 Lessons from past disruptions 29 4.2 Making coordination possible 30 4.3 Roadmap to delivery 31 Conclusion 33 Contributors 34 36 # Disclaimer This document is published by the World Economic Forum as a contribution to a project, insight area or interaction. The findings, interpretations and conclusions expressed herein are a result of a collaborative process facilitated and endorsed by the World Economic Forum but whose results do not necessarily represent the views of the World Economic Forum, nor the entirety of its Members, Partners or other stakeholders. © 2025 World Economic Forum. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, or by any information storage and retrieval system. # Foreword Kiva Allgood Managing Director, World Economic Forum Jonathan Price President and Chief Executive Officer, Teck Resources Jeremy Weir Chairman, Trafigura Group Igor Hulak Partner and Global Head of Mining and Metals, Kearney The race to electrify, digitalize and decarbonize the global economy relies on resilient, transparent and sustainable minerals and metals supply chains – building the foundation on which both today's global economy and the ongoing energy and digital transition rest. Yet these complex supply chains face increasing strain: they span continents, operate on long permitting timelines and remain highly concentrated in a small number of regions. Meanwhile, global demand centres are shifting rapidly – from the accelerating pace of industrial activity to the unprecedented build-out of electricity systems, data centres and clean-energy infrastructure. Together, these pressures create systemic vulnerabilities that can disrupt everything from infrastructure delivery to industrial competitiveness. To confront these challenges, the World Economic Forum, in collaboration with Kearney and with support from the Forum's Mining and Metals Industry Community, has developed this focused work stream under the Future-Proofing Global Value Chains initiative to build awareness and outline practical collective actions for strengthening resilience in the minerals and metals supply chains across electric vehicles (EVs), data centres and grid infrastructure. Governments, companies and investors each have a role to play in streamlining permitting, scaling secondary and circular supply, strengthening platforms for ongoing dialogue and creating shared accountability. Meaningful progress will also depend on deeper coordination with the global public-sector leaders who shape policy, infrastructure planning and long-term investment signals. Achieving this calls for structured collaboration across sectors, industries and regions at a scale not seen before. Although collaboration across regions and sectors has long existed, its nature is evolving. Today, regional alliances, cross-value chain partnerships and public-private platforms are moving beyond dialogue and compliance towards more integrated, data-driven ways of working that can drive progress at multiple levels. When connected through common standards, shared data and mutual recognition, these local and regional efforts can reinforce one another and strengthen the resilience of the overall system. Driving this type of coordinated action requires creating the conditions for actors to align on shared priorities and timelines, each within their scope. Initiatives that convene diverse stakeholders, foster knowledge exchange and align incentives can transform fragmented efforts into tangible progress. These actions will be essential as we elevate this resilience agenda across value chains. If we act together, we can future-proof global value chains and ensure that the materials essential to the energy transition and other systemic global transformations are delivered reliably, responsibly and at the scale the world demands. # Executive summary The resilience of the minerals supply chains that underpin electrification and digitalization depends on structured multistakeholder coordination. The pace and scale of electrification and digitalization depend on resilient, transparent and well-coordinated minerals and metals supply chains. Electric vehicles (EVs), data centres and electricity transmission and distribution (ET&D) each rely on the same materials, such as copper, aluminium and rare earth elements. These are processed through a small number of refining hubs, making resilience a shared priority for industry and government. Yet, the system remains only partially understood. Interconnections and vulnerabilities across value chains run deeper than many stakeholders realize. Shared refining hubs and common upstream bottlenecks mean that a disruption in one sector can reverberate across many others. These links, however, are rarely reflected in systemic or coordinated policy decisions – many recent measures address minerals supply and security in isolation rather than as part of an integrated value-chain approach. Early warning signs are already apparent: new capacity for critical minerals remains unpermitted or unfunded, while demand continues to outpace upstream investment. Against this backdrop, the World Economic Forum, in collaboration with Kearney and enabled by the support of the Forum's Mining and Metals Industry Community, consulted with more than 65 senior executives across these value chains to hear their perspectives and shape a shared understanding of the issues. These are captured in this white paper, which aims to: 1. Clarify interdependencies: Show how mineral dependencies connect sectors that were once thought of as distinct. EVs, grids and data centres rely on many of the same materials and refining and manufacturing hubs, yet decisions are still taken in silos (Sections 1-2). 2. Strengthen shared awareness: Identify where vulnerabilities lie, how technology and policy shifts shape demand and why fragmented responses amplify risk (Section 3). 3. Drive collective actions: Emphasize that coordination across supply chains, investment timelines and policy frameworks is essential to manage these risks and ensure capacity grows where and when it is needed (Section 4). Resilience will not emerge organically. It requires structured collaboration – connecting decisions, priorities and investment timelines across industries, financiers and governments. Progress will depend on empowered coordination platforms and partnerships, capable of convening diverse actors and aligning their efforts towards shared outcomes. Building on findings from the Forum's Securing Minerals for the Energy Transition (SMET) initiative, this paper outlines four key building blocks for collective action to translate collaboration into tangible progress: - Strengthen coordination mechanisms to bring together stakeholders and align priorities across sectors. Build a shared risk picture to clarify vulnerabilities, bottlenecks and interdependencies. - Deploy collective resilience strategies – combining supply diversification and expansion with demand-side measures. - Enable collaboration through convening platforms, transparency tools and common standards. This decade will be decisive for the energy and digital transformations already underway. Delivering vehicles, data capacity and grid infrastructure on time to meet the pace and scale of these transitions will depend on multistakeholder coordination - anchored in shared awareness of the mineral systems that connect them all. 1 # Understanding mineral intensity Resilient mineral supply chains will determine how fast electrification, digital infrastructure and global competitiveness can scale. Understanding how mineral reliance connects these interdependent systems is the first step towards building resilience. EVs, data centres and ET&D underpin modern economies and daily life. Each drives electrification and digitalization, relying on overlapping materials, suppliers, refining capacities and interdependent systems. Batteries, semiconductors and transformers rely on overlapping refining and manufacturing hubs concentrated in a few countries. EVs charge on the grid; data centres use the grid to operate; the grid needs a steady flow of transformers, cables and power electronics; and global value chains depend on data. Recognizing these interdependencies is essential for effective planning and coordinated action. FIGURE 1 Overview of growth trajectory, and minerals and metals used for select value chains Note: Based on International Energy Agency (IEA) announced pledges scenario Sources: S&P Global, IEA, $^{1}$ Bank of America Securities, $^{2}$ Bloomberg, Kearney analysis Mineral intensity has two aspects: scale and criticality. Sectors with a large share of global demand – such as EVs for lithium, and grids for copper and vanadium – can shape market signals, steer investment and accelerate innovation, but are also mutually exposed to vulnerabilities faced by producers that, in turn, rely on their bankable demand to scale capacity. Sectors with smaller demand have many essential specialty inputs – such as data centres’ reliance on gallium, germanium and rare earth elements (REEs) – yet have limited sway over supply or policy decisions and must track the moves of bigger buyers. Figure 2 shows why both groups need to engage: those with scale influence outcomes; those with critical need feel the consequences. FIGURE 2 Value chain demand contribution across tracked minerals (% of total global demand, 2035e) Note: ${}^{1}\mathrm{C}$ : Graphite Sources: IEA, Shanghai Metals Market, CRU, Association for Iron & Steel Technology, International Lead and Zinc Study Group, United States Geological Survey, Silver Institute, Research and Markets, International Tin Association, S&P Global and Kearney analysis The next three subsections map where minerals sit inside each product system. # 1.1 The known frontier of electric mobility Industry consultations and extensive literature review reflect broad awareness that EVs are highly mineral-intensive. By 2035, EVs are projected to account for $86\%$ of total lithium demand, $55\%$ of cobalt and one-third of global rare-earth consumption, placing their supply chains at the centre of future mineral balances. The following breakdown illustrates how key EV components translate into distinct minerals' demand, each shaping exposure and supply dependencies differently. # Battery pack Batteries are a key driver of EV mineral demand. Two cathode chemistries dominate: nickel-manganese-cobalt (NMC) and lithium-iron-phosphate (LFP). Chemistry choices shape EV cost and performance and the geographic footprint of upstream refining and processing, with the shift towards LFP reducing reliance on cobalt and nickel. # Motor Traction motors and associated power electronics, while relatively small in mass, add another layer of material exposure. Permanent-magnet motors rely on REEs such as neodymium (Nd) and dysprosium (Dy), alongside copper in windings and busbars. # Wiring and power electronics Power electronics and wiring make EVs significantly more copper-intensive than internal combustion engine (ICE) vehicles, with aluminium substitutions in some designs where standards allow and dependence on silicon and emerging silicon-carbide (SiC) semiconductors, whose specialized manufacturing capacity is tight against rising demand. # Structure (chassis and body) Metal mass in EVs is dominated by steel and aluminium. These are mature, globally traded metals with established recycling flows, but their scale makes them production-critical. Trend: Evolving battery chemistries, motor designs and power electronics are reshaping material demand across EV value chains – reducing reliance on cobalt and nickel through LFP adoption, raising copper and semiconductor needs through electrification of drivetrains and increasing aluminium use through light-weighting and megacast structures.3 Total minerals and metals mass in passenger EVs: $\sim 900 - 1000\mathrm{kg}$ # EV wiring and power electronics $\sim 3 - 4\%$ of minerals and metals used in EV Copper \~27 Aluminium \~3 Gold <0.1 Silicon <0.1 Silver <0.1 $\bullet$ Tin \~0.2 Tantalum <0.1 # EV motor $\sim 2 - 4\%$ of minerals and metals used in EV Rare earths \~0.7 Steel \~20 Copper \~7 Aluminium \~1.4 # Structure (chassis and body) $\sim 50 - 65\%$ of minerals and metals used in EV Rare earths \~385 Steel \~145 Copper \~8 Aluminium \~12 # EV battery pack $\sim 30 - 40\%$ of minerals and metals used in EV Graphite \~40 Silicon \~1 Aluminium \~30 Copper \~20 Steel \~20 <table><tr><td></td><td>Cathode chemistry</td><td>High nickel</td><td>Low/mid nickel</td><td>LFP</td><td>High manganese</td><td>Sodium-ion</td></tr><tr><td>● Lithium</td><td>~8</td><td>~8</td><td>~8</td><td>~8</td><td>~8</td><td></td></tr><tr><td>● Cobalt</td><td>~5</td><td>~10</td><td></td><td>~5</td><td>~20</td><td></td></tr><tr><td>● Nickel</td><td>~40</td><td>~30</td><td></td><td>~20</td><td></td><td></td></tr><tr><td>● Manganese</td><td>~5</td><td>~15</td><td></td><td>~80</td><td>~20</td><td></td></tr><tr><td>● Iron</td><td></td><td></td><td>~40</td><td></td><td></td><td>~25</td></tr><tr><td>● Phosphorus</td><td></td><td></td><td>~20</td><td></td><td></td><td>~10</td></tr></table> # Plug-in hybrid EV (leveraging lead-acid battery) $\bullet$ Lead \~12 Value chain contribution to global demand of minerals and metals (2035) Note: Based on average EV battery size in 2024 of 60 kilowatt-hour (kWh); percentage value refers only to the mass of metals and minerals and not the total vehicle mass. Sources: S&P Global, ICCT, IEA and Kearney analysis # 1.2 The hidden footprint of digitalization Consultations revealed that awareness of mineral reliance among data centre actors remains low despite accelerating demand. Rapid expansion of hyperscale and AI-optimized facilities is creating a new layer of mineral demand beyond traditional manufacturing. By 2035, data centres are projected to account for about $6\%$ of global gallium use and some $2.4\%$ of germanium demand. While relatively small in volume, these minerals are critical to enable efficient, high-performance computing. Breaking down the data centre into its key subsystems highlights how different functions – from computing to cooling – depend on distinct sets of materials and drive exposure to a wide range of refined metals. # Servers and chips Although moderate in overall mass, the compute layer concentrates high-value, high-purity materials – silicon, gallium, germanium, gold, silver and tin – that underpin performance and reliability. # Networking and storage Networking and facility-level storage hardware represents limited mass but depends on a complex mix of refined metals (such as copper and aluminium) and specialty materials to deliver conductivity, signal precision and thermal stability. # Cooling systems Cooling and heat-rejection infrastructure accounts for the largest share of total metal mass, dominated by aluminium, copper and steel. # Electrical systems Power-distribution - switchgear, transformer cabling - is among the most copper- and iron-intensive part of data centre construction and a major cost and schedule driver. # Backup power Backup systems integrate generators, batteries, converters and switchgear that collectively anchor power reliability and build system redundancy. Trend: Rising compute intensity, power demand and thermal loads are reshaping materials use in data centres – increasing reliance on copper and aluminium for power and cooling systems, expanding exposure to gallium, silicon carbide and other specialized semiconductors, and adding new dependencies on lithium-ion technologies for backup power.<sup>6</sup> Total minerals and metals mass per MW: $\sim 60 - 75$ metric tons # Networking and storage $\sim 1 - 3\%$ of minerals and metals used in data centres # Network hardware and facility-level storage Aluminium \~260 Cobalt \~1 Copper \~280 Germanium \~0.1 Gallium \~0.3 Gold <0.1 Graphite \~1 $\bullet$ Iron \~660 Nickel \~15 $\bullet$ REEs \~1 Silicon \~20 Silver \~0.5 $\bullet$ Tin \~10 # Servers and chips (compute $\sim 12 - 16\%$ of minerals and metals used in data centres # CPUs, GPUs, memory, local storage Aluminium \~2,700 Cobalt \~2 Copper \~2,000 Germanium \~0.3 Gallium \~2 Gold \~1 Graphite \~5,100 $\bullet$ Iron \~15 Lead \~35 Nickel \~2 REEs ~160 Silicon \~5 Silver \~5 $\bullet$ Tin \~120 # Electrical systems $\sim 15 - 20\%$ of minerals and metals used in data centres # Switchgear, PDUs, $^{1}$ cabling Aluminium \~1,200 Copper \~40 Germanium \~2,200 Gold <0.1 $\bullet$ Iron \~290 Nickel \~14,000 $\bullet$ REEs \~240 Silcon \~30 Silver \~220 $\bullet$ Tin \~0.2 # Cooling systems $\sim 35 - 45\%$ of minerals and metals used in data centres # CRAH/C, CDUs, chillers, cooling tower Aluminium \~6,700 Copper \~3,600 Gold <0.1 Graphite \~15 $\bullet$ Iron \~18,000 Lead \~10 Nickel \~130 REEs <0.1 Silicon \~15 Silver <0.1 $\bullet$ Tin \~0.1 # Backup power $\sim 20 - 30\%$ of minerals and metals used in data centres # Generators, UPS (li-ion/lead-acid) Aluminium \~1,200 Cobalt \~40 Copper \~2,200 Gold <0.1 Graphite \~290 Iron \~14,000 Lead \~240 Lithium \~30 Nickel \~220 $\bullet$ REEs \~0.2 Silicon \~10 Silver \~0.5 $\bullet$ Tin \~2 Value chain contribution to global demand of minerals and metals (2035) Notes: Based on a 2024 split of conventional (CPU or central processing unit-driven) data centres at $76\%$ capacity vs. accelerated (GPU or graphics processing unit-driven) data centres at $24\%$ capacity. $^{1}$ Power distribution unit. $^{2}$ Computer room air handler/conditioner. $^{3}$ Cooling distribution unit. Source: Bank of America Securities, IEA and Kearney analysis # 1.3 Copper and steel for electrification Awareness of the grid's mineral intensity – particularly its reliance on copper and grain-oriented electrical steel (GOES) – is rising, as reflected by recent studies and by the consultations with industry players. ET&D systems rank among the most metal-intensive infrastructure segments, comparable to transport and construction in their reliance on bulk metals.[7] By 2035, grids are expected to account for about $29\%$ of global vanadium demand, $18\%$ of copper and $7\%$ of lithium, underscoring how expansion will increasingly depend on secure, diversified mineral supply. Decomposing the grid into its main physical components shows where material demand concentrates and how it defines project timelines and resilience. # Towers and poles Structures account for the largest share of total ET&D metal mass, driven by extensive use of steel, aluminium and protective zinc coatings. Steel fabrication determines both cost and lead time, while aluminium reduces weight for long-span transmission lines. # Substations and transformers Substations and transformers combine copper windings, GOES cores, and structural iron and aluminium frames. # Electricity cables Copper and aluminium are the principal conductors in transmission and distribution cables, supported by steel for sheathing and structural reinforcement. Material selection depends on voltage, cost and weight considerations. # Battery energy storage systems (BESS) BESS are emerging as a new grid asset class to reinforce the reliability of fluctuating renewable power generation. Material needs vary by chemistry: lithium-ion systems rely on lithium, nickel and graphite; sodium-ion systems use iron and aluminium; and vanadium redox-flow systems draw on vanadium and phosphorus. Trend: Grid expansion and electrification are intensifying demand for copper, aluminium and electrical steel, and tightening supply chains for cables and transformers, while emerging battery storage technologies add new dependencies on lithium, iron, phosphate and vanadium.8 Total minerals and metals mass per km ET&D grid length (2035): $\sim 9,000 - 12,000\mathrm{kg}$ # Towers and poles $\sim 40 - 50\%$ of minerals and metals used in ET&D Minerals Poles Towers Aluminium \~510 \~450 Zinc ~102 50 - Iron $\sim 1,200$ $\sim 2,700$ # Battery energy storage $\sim 5 - 10\%$ of minerals and metals used in ET&D Minerals Transmission Distribution <table><tr><td>Minerals</td><td>Redox flow</td><td>Sodium ion</td><td>LFP</td></tr><tr><td>Vanadium</td><td>~700</td><td></td><td></td></tr><tr><td>Graphite</td><td>~20</td><td></td><td>~200</td></tr><tr><td>Iron</td><td>~1,500</td><td>~200</td><td>~150</td></tr><tr><td>Aluminium</td><td>~600</td><td>~60</td><td>~150</td></tr><tr><td>Copper</td><td>~80</td><td></td><td>~100</td></tr><tr><td>Phosphorus</td><td></td><td>~300</td><td>~70</td></tr><tr><td>Lithium</td><td></td><td></td><td>~20</td></tr><tr><td>Silicon</td><td></td><td></td><td>~100</td></tr></table> # Substations and transformers $\sim 25 - 27\%$ of minerals and metals used in ET&D Transformers, switchgear and circuit breakers, busbars and connectors, yard structures Aluminium \~6,700 Electrical steel \~3,600 Copper <0.1 - Iron \~15 Zinc \~18,000 # Electricity cables $\sim 20 - 30\%$ of minerals and metals used in ET&D Transmission/distribution lines (overhead and underground) Aluminium \~1,200 700\~1,700 Copper \~500\~2,500 - Iron $\sim 500$ $\sim 400$ Value chain contribution to global demand of minerals and metals (2035) Notes: Figures are normalized to kg/km: lines from typical conductor sizes and spans; substations/transformers apportioned by asset density per km; civil works excluded Sources: Bloomberg, IEA, OEMs and Kearney analysis # Supply chains disjointed by tiers, territories and timelines Multi-tiered and geographically dispersed value chains grow differently, creating complex interdependencies and testing coordination and system resilience. EVs dominate battery-metal consumption, the grid takes up the bulk of metals like copper and aluminium, and data centres need high-purity inputs for chips and power electronics. Together, they set the tone for global minerals trade and processing priorities. The three value chains share a common structure that begins with mineral extraction and extends through multiple processing and manufacturing stages to final assembly and operation. Each tier involves different stakeholders, from miners and chemical processors to component makers, original equipment manufacturers (OEMs), engineering, procurement and construction companies (EPCs) and utilities, often operating across distant geographies, which adds to the overall complexity and interdependence. This complexity, amplified by the distance between tiers, creates both awareness and visibility gaps – making it difficult for stakeholders to anticipate where capacity will be needed or how timelines align, underscoring the importance of early coordination and planning for resilience. Consultations across upstream and downstream participants confirmed that they don't communicate frequently. Many of the downstream players consulted said they do not yet track minerals risks as continuity risks and almost all pointed to the "distance" between tiers as a barrier to coordination. Across value chains, capacity additions follow similar patterns but on very different timelines. - Mining projects in some jurisdictions require 10-20 years from discovery to production, reflecting permitting, infrastructure and financing hurdles.[9] Refining facilities take three to eight years $^{10}$ to develop, depending on permitting requirements, construction complexity and customer qualification. - Manufacturing and end-product facilities can be commissioned within one to five years, reflecting shorter construction cycles and fewer regulatory constraints. These disparities matter most when new capacity is urgently needed, as downstream segments can scale far faster than upstream supply can adjust - impacting the underlying demand drivers for materials. These indicative lead times are common across EVs, data centres and ET&D, but each value chain exhibits specific bottlenecks and timing gaps that will be discussed in the following subsections. The vulnerability doesn't arise from linear queues, it arises when a specific enabler - permits, technology and supplier qualification, component capacity or interconnections - lags behind demand. There is also a strategic planning horizon mismatch – miners plan for 10- to 15-year cycles while downstream players typically look two to five years ahead. Misaligned development timelines create a structural challenge for supply resilience. When upstream investment slows or permitting slips – or when a mine sanctioned on yesterday's outlook comes online after technology, specifications or siting have shifted – projects face higher costs, re-sourcing/retooling or idle capacity that cascades downstream. Such timing mismatches amplify costs and delays across value chains. Anticipating these timing gaps and planning capacity early is essential to keep supply and demand aligned as electrification and digitalization accelerate. The following sections examine how these dynamics manifest across EV, data centres and grid value chains - each revealing distinct pressures, geographic realities and opportunities for resilience. # 2.1 | Anatomy of the electric mobility supply chain A tightly integrated yet geographically concentrated system defines EV manufacturing. Battery, motor and electronics production have scaled rapidly, but dependence on a few refining and assembly hubs leaves the chain exposed to policy, trade and logistics risks. Supply tiers that multiply exposure: Each stage of the EV value chain involves more than 10 distinct tiers of suppliers, from miners and chemical processors to cathode, anode and component manufacturers, creating interdependent lead times across the system. - Fast ramp-up, slow foundations: New EV manufacturing assets – battery gigafactories, $^{11}$ motors, power electronics and structural components – typically ramp up within one to three years, $^{12}$ with final assembly over two years, $^{13}$ far faster than mining or refining capacity can adjust. - Circularity still waiting to kick in: End-of-life management remains nascent. Recoverable battery volumes are expected only after 2030, as vehicles sold in the early 2020s reach retirement age.[14] - Geography as both advantage and exposure: The industry's production footprint is highly concentrated. Battery-cell manufacturing and the underlying cathode/ anode technology and R&D are used as a proxy for manufacturing locations. China holds the dominant market shares: approximately $50\%$ of global minerals refining, $70\%$ of battery manufacture and $70\%$ of EV assembly.[15] This structure has enabled rapid cost declines through integrated supply chains and learning curve effects, but it also heightens exposure to policy, trade and logistics disruptions that can cascade through the value chain.[16] Key: Market share by geography (% of total value) China Asia Europe North Latin America Middle East Africa Australia Other Sources: USGS, ${}^{17}$ IEA and Kearney analysis # 2.2 Structure of the digital infrastructure supply chain An increasingly material-intensive digital backbone faces limited visibility across a long, complex chain. Rapid growth in hyperscale and AI-optimized facilities has multiplied dependencies on high-purity metals and semiconductors, but awareness of this reliance remains low among many operators. - Interconnected systems, invisible dependencies: The data centre value chain links upstream materials and component manufacturing to complex facility integration. Multiple subsystems – computing, power, cooling and networking – are assembled in parallel, creating more than 14 interdependent supply tiers. Timelines that can't keep up with compute demand: Greenfield semiconductor fabrication plants typically require three to five years from investment to output,[18] while servers,[19] power and cooling[20] equipment facilities are completed in one to three. New-build data centres take nine to 36 months, often in phases depending on scale. - Circularity – still more potential than practice: End-of-life and secondary supply from decommissioned facilities yield recoverable copper, aluminium, steel and REEs. Circular practices for semiconductors and unlimited power supply (UPS) batteries remain limited but are expected to scale later in the decade. - Concentration where chips are made – and consequent risks: Semiconductor manufacturing – central to CPUs, GPUs and memory chips – remains concentrated in East Asia (China has a $24\%$ market share, Taiwan $18\%$ , South Korea $18\%$ and Japan $15\%$ ).<sup>21</sup> Construction is distributed across North America and Europe, and emerging in Asia. This split supports cost efficiency but increases exposure to semiconductor bottlenecks, trade restrictions and logistics risks. Note: ${}^{1}$ Battery management system Sources: USGS, IEA, JP Morgan Asset Management, Gartner and Kearney analysis # 2.3 Map of the grid supply chain Grid infrastructure combines long build cycles with uneven regional capacity. Expansion depends on heavy industrial inputs and sequential processes that define both timing and resilience. - Complex supply tiers that dictate the pace of progress: Multiple industrial tiers (more than 12) – conductors, transformers, switchgear and structural components – interact through long build-out sequences that set the pace for grid expansion. - Lead times measured in years, not quarters: New transformer factories take two to four years to commission,[22] high-voltage direct current (HVDC) facilities three to five,[23] and tower/ pole fabrication plants 12 to 24 months.[24] Transmission-line and substation integration extends nine to 36 months, depending on terrain and coordination. - Circularity limited by asset longevity: Unlike vehicles or electronics, grid assets have decades-long lifespans and slow turnover. Near-term secondary flows are therefore modest, though recycling of decommissioned copper and steel components is well-established. - Geography shapes supply – and delays: Manufacturing capacity is regionally dispersed but uneven. China and East Asia account for roughly $31\%$ of transformer output, supported by large domestic steel and aluminium industries. Europe and North America retain advanced design and testing capabilities but face shortages of large-power transformers and HVDC equipment, extending delivery times and delaying grid modernization. Sources: USGS, Transformers Magazine,25 IEA and Kearney analysis 3 # Shared risks, divided attention Minerals-intensive value chains face physical bottlenecks and systemic gaps that can quickly delay equipment deliveries and increase project costs for downstream players. Mineral supply for EVs, data centres and ET&D faces two kinds of exposure: physical bottlenecks and system frictions. Concentrated processing, long build times and site constraints limit how fast supply can grow, while weak cross-tier visibility, shifting demand and fragmented standards slow investment and qualification. Because mining, refining and equipment manufacturing operate on different timelines and face distinct vulnerabilities, risk management remains fragmented, turning small shocks into downstream delays and higher costs. Additionally, vertical integration rarely extends beyond one to three tiers, leaving miners, refiners, component makers and end-product owners weakly connected at the system level. These findings build on the Forum's Securing Minerals for the Energy Transition work,[26] which has identified supply-demand gaps and the societal, environmental and governance risks that continue to shape resilience today. Despite differing perspectives shared during consultations, four systemic themes have emerged: - Fragmented dialogue and visibility gaps: Distance between upstream and downstream actors creates information gaps and an uneven understanding of one another's risks; dialogue is episodic and bilateral rather than system-wide. - Demand uncertainty: Unclear and constantly changing demand projections across the value chain, driven partly by policy changes and rapidly evolving technologies, hinder investments and planning. - Differing risk agendas: Each tier in the value chain optimizes for its own risks and targets, creating misaligned incentives across the value chain. - Low salience and actionability gap: Mineral risks often seem marginal and sit below nearer-term issues (energy costs, emissions, tier-1 performance); even when flagged, downstream players lack clear, practical levers to act, leaving issues unresolved. EV OEMs are further along in managing upstream risk than data centres and many ET&D players. The semiconductor shortage halted vehicle production and exposed the limits of single-tier visibility. In response, leading EV manufacturers elevated mineral risk to the executive agenda to avoid repetition. By contrast, mineral governance is less mature in data centres and grids, though some utilities already maintain strong supplier visibility. # 3.1 | Early warning signs in supply This section focuses on the minerals and metals that drive the bills of materials for EV, data centre and ET&D value chains, examining capacity signals, supply-demand balances and sources of supply # Emerging supply volatility Overall supply-demand balances in 2024 were stable, with small gaps in select materials, such as manganese and aluminium, reflecting acute market dynamics – temporary shutdowns, rapid price swings, logistical constraints – rather than structural shortfalls. However, supply volatility is already visible. In Yunnan, China, droughts repeatedly cut hydropower, curbing aluminium output.[27] Shutdowns like at Bald Hill in Australia removed lithium supply and exposed market fragility.[28] Smelter closures and cutbacks reflect squeezed economics from low – and even negative – copper treatment and refining charges (TC/RCs), with governments and industry warning current fee levels are unsustainable.[29] Growing regionalization and policy actions are re-routing flows and reinforcing concentration: China introduced export controls on gallium and germanium in $2023^{30}$ and later restricted rare-earth technologies and material exports;<sup>31</sup> Indonesia's nickel-ore export ban increased and concentrated nickel-processing capacity within the country;<sup>32</sup> and the Democratic Republic of the Congo (DRC) imposed cobalt export restrictions,<sup>33</sup> further tightening global supply chains. Even without shocks, output from the current asset base erodes as reserves deplete and ore grades decline.<sup>34</sup> In copper, for example, maintaining flat supply requires ongoing investment – new and sustaining projects must first replace lost volumes before adding net new capacity. At the same time, new Western projects are advancing. Keliber's lithium hydroxide project in Finland was designated a "Strategic Project" under the European Union's Critical Raw Materials Act (CRMA), a status intended to accelerate permitting and enable access to finance.[35] Thacker Pass secured a $2.26 billion US Department of Energy loan,[36] and the US is channelling cross-border support to allied Canadian projects under the Canada-US Joint Action Plan on Critical Minerals Collaboration to strengthen global supply-chain resilience. The next decade remains a window to prepare before steeper demand tests capacity. Looking to 2035, forecast tightness stems not only from demand growth but from how much of 2035 supply still needs to be developed. Figure 9 projects that for several minerals, today's operating base covers only half (or less) of expected 2035 demand, with the balance dependent on projects that must still be permitted, financed or even identified. For example, today's supply of lithium and graphite covers approximately $35 - 45\%$ of forecasted 2035 demand; even if announced projects deliver, material shortfalls are expected to persist – implying a near-doubling of output needed within a decade. Recycling helps but cannot close these gaps alone. Its contributions are meaningful for copper and aluminium – the International Energy Agency (IEA) estimates an average recycled-input share of about $35\%$ for aluminium and approximately $17\%$ for copper (excluding direct-use scrap) over the last decade.[37] However, recycling of lithium and rare earths remains nascent due to low end-of-life volumes and insufficient collection and mechanical-separation capacity. FIGURE 9 Projected global supply-demand balance for select minerals (2035e)1 Notes: Balances are indicative and synthesized from multiple sources based on available information; figures may vary by source and assumptions (e.g. demand trajectories, project start-up/ramp, recycling rates). 1 Percentage shortfall/surplus = (forecast refined supply - forecast refined end-demand) ÷ forecast demand (2035e). 2 C: Graphite 3 PO: Phosphates Sources: IEA, Shanghai Metals Market, CRU (Commodities Research Unit), Association for Iron & Steel Technology, International Lead and Zinc Study Group, USGS, Silver Institute, Research and Markets, International Tin Association, International Aluminium Institute, S&P Global, Wood Mackenzie and Kearney analysis # Concentrated sources of supply Across EVs, data centres and ET&D, mining is relatively more dispersed, but refining and processing are highly concentrated. The IEA estimates that the top three refining countries' average market share rose from approximately $82\%$ (2020) to $86\%$ (2024),<sup>38</sup> a level of concentration rare in other industrial inputs. FIGURE 10 Countries with $\geq 5\%$ share of mined and refined production of select minerals (2024) Notes: 1 Includes countries that are involved in mining and/or refining but do not constitute $\geq 5\%$ market share for the tracked minerals. Sources: IEA, S&P, USGS and Kearney analysis Mining: Of the $19^{39}$ tracked minerals, China has $\geq 5\%$ mined share in 15, Australia in seven and Brazil and Russia each in six - still concentrated for several minerals but more geographically spread than refining. Refining: China holds $\geq 5\%$ of refined output for all 19 tracked minerals and exceeds $75\%$ for several (gallium, graphite, REEs, magnesium, germanium and cobalt). Exceptions underscore the rule: Nickel processing is now dominated by Indonesia (often via Chinese-owned assets) and gold refining is more distributed, with Switzerland a major hub. Other smaller hubs include India with market shares $\geq 5\%$ in four bulk metals (aluminium, iron, zinc and lead) and Japan with four (tantalum, copper, gold and silver). <table><tr><td>Minerals</td><td colspan="3">Reserve life index (years)</td><td colspan="4">Mining and extraction Largest miner and share of top 3 countries (%)</td><td colspan="3">Refining and processing Largest refiner and share of top 3 countries (%)</td></tr><tr><td>Gallium1</td><td colspan="3">65</td><td colspan="3">Guinea</td><td>72%</td><td colspan="2">China</td><td>100%</td></tr><tr><td>Graphite</td><td colspan="3">170</td><td colspan="3">China</td><td>88%</td><td colspan="2">China</td><td>98%</td></tr><tr><td>REEs</td><td colspan="3">230</td><td colspan="3">China</td><td>91%</td><td colspan="2">China</td><td>97%</td></tr><tr><td>Germanium</td><td colspan="3">420</td><td colspan="3">China</td><td>79%</td><td colspan="2">China</td><td>96%</td></tr><tr><td>Lithium</td><td colspan="3">125</td><td colspan="3">Chile</td><td>94%</td><td colspan="2">China</td><td>96%</td></tr><tr><td>Magnesium</td><td colspan="3">350</td><td colspan="3">China</td><td>78%</td><td colspan="2">China</td><td>94%</td></tr><tr><td>Cobalt</td><td colspan="3">40</td><td colspan="3">Democratic Republic of Congo</td><td>88%</td><td colspan="2">China</td><td>89%</td></tr><tr><td>Nickel</td><td colspan="3">40</td><td colspan="3">Indonesia</td><td>75%</td><td colspan="2">Indonesia</td><td>78%</td></tr><tr><td>Iron</td><td colspan="3">80</td><td colspan="3">Australia</td><td>66%</td><td colspan="2">China</td><td>77%</td></tr><tr><td>Tin</td><td colspan="3">15</td><td colspan="3">China</td><td>52%</td><td colspan="2">China</td><td>75%</td></tr><tr><td>Manganese</td><td colspan="3">90</td><td colspan="3">South Africa</td><td>80%</td><td colspan="2">China</td><td>73%</td></tr><tr><td>Aluminium</td><td colspan="3">65</td><td colspan="3">Guinea</td><td>72%</td><td colspan="2">China</td><td>71%</td></tr><tr><td>Tantalum</td><td colspan="3">1902</td><td colspan="3">Democratic Republic of Congo</td><td>78%</td><td colspan="2">China</td><td>66%</td></tr><tr><td>Zinc</td><td colspan="3">20</td><td colspan="3">China</td><td>54%</td><td colspan="2">China</td><td>61%</td></tr><tr><td>Copper</td><td colspan="3">40</td><td colspan="3">Chile</td><td>49%</td><td colspan="2">China</td><td>57%</td></tr><tr><td>Gold</td><td colspan="3">20</td><td colspan="3">China</td><td>30%</td><td colspan="2">Switzerland</td><td>57%</td></tr><tr><td>Lead</td><td colspan="3">20</td><td colspan="3">China</td><td>61%</td><td colspan="2">China</td><td>56%</td></tr><tr><td>Phosphates</td><td colspan="3">300</td><td colspan="3">China</td><td>67%</td><td colspan="2">China</td><td>55%</td></tr><tr><td>Silver</td><td colspan="3">30</td><td colspan="3">Mexico</td><td>51%</td><td colspan="2">China</td><td>43%</td></tr></table> Notes: ${}^{1}$ Gallium derives mainly from bauxite (aluminium) ore and not as standalone deposits; mining concentration reflects bauxite production Tantalum rse i t a Australia, Brazil and China Sources: IEA, S&P, USGS and Kearney analysis Short reserve-life combined with high production and refining concentration elevates risk exposure. Bulk system metals such as copper warrant special attention given their centrality across all three value chains. Additionally, for several specialty materials – graphite, REEs, germanium, lithium, magnesium, cobalt, nickel, manganese and tantalum – the top three mining and refining countries hold $>75\%$ of the market share. This heightens exposure to single-point disruptions such as operational outages or export controls, which can trigger global price or schedule shocks – even when aggregate supply appears adequate. Ownership patterns add another layer: cross-border corporate control can diversify financing and expertise, but it can also transmit policy risk across regions, thereby shifting strategic control of assets. For the minerals tracked in this report, a disruption at the largest producer creates a material risk of unmet global demand and lasting downstream impacts, making market concentration a key challenge to supply chain resilience. # 3.2 | Upstream shocks, downstream consequences Because processing is concentrated and qualification of new suppliers, materials or technologies takes time, local shocks upstream quickly become delivery risks downstream. Figure 12 maps the chain of disruption, outlining how issues in mining and refining can translate into component delays and end-product slippage. One example is GOES, essential for transformer cores, where limited global capacity – further squeezed by the Russia/Ukraine-related trade restrictions – has collided with strong demand and skilled-labour shortages. With GOES scarce, transformer prices have increased 50-80% over the last three years and lead times have stretched to about 18-36 months, forcing order reprioritization.40 Downstream, utilities and developers have delayed grid connections, pushing back EV-charger rollouts and data centre energization. Price dynamics matter. Spikes can unlock financing and accelerate projects; beyond a threshold, they trigger substitution and design pivots, increasing uncertainty for suppliers. In such periods, investors may pause or delay final investment decisions (FIDs) until technology choices and demand signals stabilize, prolonging the bottlenecks that higher prices were meant to relieve. FIGURE 12 Cascading risk impact from upstream production to downstream delivery Across very different bottlenecks, the risk pattern is similar: capacity may exist in aggregate but arrives too late or at a higher price than anticipated, so costs rise and schedules slip before supply can be requalified or redirected. Demand surges in one value chain tighten the same metals and equipment others need; as EV, data centre and grid build-outs draw on overlapping inputs such as copper, electrical steel and semiconductors, short-term trade-offs emerge – reinforcing the need for dialogue and coordinated planning. With capacity concentrated in a few hubs, policy or operational shocks propagate across the system. Without coordination, procurement becomes zero-sum – with higher prices and longer delays for everyone. # 3.3 Competing risk agendas Risk perspectives differ by tier and shared understanding of each other's exposure is uneven, yet recurring patterns appear across EVs, data centres and ET&D, compounding from mining to end products. During consultations, executives across value chains shared their top risks, which were subsequently translated into seven risk vectors. Figure 13 maps how these risks concentrate across tiers and highlights how priorities differ by value chain step. High supply concentration amplifies each, turning local shocks into global impact. Across consultations, three risks dominated: - Geopolitics and trade: Proliferating trade rules that regionalize supply and increase lead-time and price uncertainty. Market and financing: Industrial policy volatility - incentives, local-content rules, licensing - that compress margins and rewire demand signals. Infrastructure: Not just ports, rail and power to upstream sites, but thin capacity for critical components and downstream grid readiness that gate new capacity and extend queues. FIGURE 13 Risk perspectives across the value chain Note: The heatmap reflects the frequency with which participants cited the particular risk among top risks; this is not a ranked survey. ${}^{1}$ TC/RCs: Treatment and refining charges. ${}^{2}$ GOES: Grain-oriented electrical steel Source:Authors # Geopolitics and trade Mining and refining: Trade and industrial policies are reshaping where value is added and who gets material first. Export controls (e.g. on REEs, gallium and germanium), sanctions and tariffs – from US duties on aluminium and copper to Canada's licence recissions $^{41}$ – have introduced new layers of planning uncertainty. Maintaining open and predictable trade remains critical to resilience and affordability. Component manufacturing and end products: OEMs in EVs and grid equipment reported heavy dependence on Chinese processing and components, leaving them vulnerable to regulatory shifts and logistics shocks. Maritime disruptions, such as the 2023-24 Red Sea $^{42}$ re-routing, demonstrated how choke points can ripple through global supply chains. Cross-border grid interconnections also face misaligned national policies and permitting delays. # Market and financing Mining and refining: Margins are tightening and investment decisions are delayed. Low TC/RCs (treatment charges or refining charges – the fees paid to smelters/refiners by miners for converting ore or concentrates) and policy volatility slow new refining projects while accelerating closures. Subsidy-driven competition, price swings and localization rules create uncertainty over where capacity will emerge – and which assets may be stranded. Component manufacturing and end products: EV suppliers cited policy volatility as a driver of demand uncertainty. Data centre component makers have visibility of only six to nine months, while utilities face procurement cycles misaligned with decade-long grid investments, limiting their ability to finance expansions. # Infrastructure Mining and refining: Access to logistics and utilities is now the critical path for new projects. Multi-billion dollar enabling infrastructure – railways, ports, desalination or power – is often required before production begins. Projects such as the QB2 copper mine expansion $^{43}$ in Chile or Simandou $^{44}$ in Guinea illustrate how the scope of infrastructure can rival that of the mine itself. Component manufacturing and end products: Transformer and HV-cable bottlenecks can hold up entire projects. Lead times of one to four years in some markets means EV chargers, data centres and grid upgrades compete for the same equipment. # Resources and energy Mining and refining: Declining ore grades and tightening power supply raises costs and emissions. Copper grades have