Future material demand for automotive lithium-based batteries

This study addresses how the transition to electric vehicles (EVs) will affect global demand for critical lithium-ion battery (LIB) materials through 2050. Against the backdrop of rapid EV adoption and concerns about sustainable supply of lithium (Li), cobalt (Co), nickel (Ni), and other materials, the objective is to quantify future material requirements under alternative EV fleet growth trajectories, battery chemistry pathways, and end-of-life (EoL) management strategies (recycling and second-use). The research aims to inform policy and industry on potential supply risks, reserve sufficiency, and leverage points—such as battery chemistry selection, improvements in specific energy, and recycling—to mitigate primary material demand and associated environmental and social impacts.

Prior work has quantified EV battery material demand for specific regions (e.g., Europe, United States, China) and for individual materials, and identified risks from concentrated supply (especially Co), environmental and social mining impacts, and uncertain reserve sufficiency. Global-level assessments (e.g., Weil et al.) indicate potential shortages for Li and Co but often do not explore the implications of evolving battery chemistries (shift to high-Ni NCM/NCA, LFP adoption, or post-Li chemistries like Li–S and Li–Air), nor systematically compare recycling and second-use effects across scenarios. This study fills these gaps by combining global EV fleet scenarios with detailed battery chemistry evolution and multiple recycling pathways to assess implications for Li, Co, Ni, Mn, Al, Cu, graphite, and Si demand.

The authors develop a dynamic, stock-driven material flow analysis (MFA) model with three layers: (1) EV fleet, (2) battery chemistries, and (3) materials. EV fleet projections to 2030 follow IEA Stated Policies (STEP) and Sustainable Development (SD) scenarios, then extrapolate to 2050 with logistic growth targeting 25% (STEP) and 50% (SD) penetration of EVs in the global light-duty vehicle (LDV) fleet. Global LDV stock grows linearly from 503 million (2019) to 3.9 billion (2050). BEV/PHEV shares evolve using EIA trends. Vehicle classes (small, mid-size, large) are assigned sales shares and performance parameters (range, efficiency, motor power) to determine average battery capacities (e.g., BEV 66 kWh, PHEV 12 kWh). EV and battery lifespans use Weibull distributions; pre-2020 batteries are assumed to last 8 years with a 50% replacement rate; after 2020, battery lifespans align with EVs (average ~15 years), with LFP assumed at 20 years. Battery chemistry scenarios: (a) NCX (NCA and NCM family with progressive shift from NCM111 to NCM523/622/811 and introduction of NCM955 by 2030; some graphite anodes include Si); (b) LFP scenario with LFP increasing to 60% market share from 2030–2050; (c) Li-S/Air scenario with Li–S and Li–Air entering in 2030 and jointly reaching 60% by 2040 onward. Pack-level specific energies used include: NCM111 ~160 Wh/kg, NCM955-Graphite(Si) ~202 Wh/kg, LFP ~129 Wh/kg, Li–S ~308 Wh/kg, Li–Air ~383 Wh/kg for mid-size BEVs. Material intensities are computed using Argonne’s BatPaC v3.1 for 48 battery variants (2 EV types × 3 size segments × 8 chemistries), with stoichiometric adjustments for chemistries not native to BatPaC. Li–S/Li–Air compositions derive from literature and are scaled to required capacities; pack component ratios mirror NCA packs. EoL management includes three recycling routes for NCX/LFP—pyrometallurgical (pyro), hydrometallurgical (hydro), and direct recycling—and mechanical recovery for Li–S/Li–Air (recovering Li metal). Material recovery efficiencies follow Argonne’s EverBatt. Economic recovery differences (e.g., Li recovery economics, Al downcycling, graphite losses in pyro) are captured qualitatively and via recovered-material assumptions. Second-use is modeled as a 10-year lifetime in stationary applications for 100% of LFP and 50% (pre-2020) rising to 75% (2020–2050) of NCX/Li–S/Li–Air, delaying recycling flows. Sensitivity analyses explore: reduced battery lifespan (continued 1.5 packs/EV after 2020), extreme battery capacities (10 kWh all-PHEV vs 110 kWh all-large BEV), 100% penetration of Co/Ni-free chemistries (LFP; Li–S/Li–Air), and higher specific energies for Li–S (600 Wh/kg) and Li–Air (1000 Wh/kg). Outputs include annual and cumulative primary material demand (2020–2050), EoL-material availability, and closed-loop recycling potential (CLRP) by material and scenario. Demands are compared to current production capacities and known reserves.

• EV fleet growth: By 2050, EV penetration reaches 25% (STEP) and 50% (SD) of the LDV fleet. EV stock increases ~72× to nearly 1 billion (STEP) and ~102× to ~2 billion (SD); annual EV sales reach ~109 million (STEP) and ~211 million (SD).
• Battery capacity needs: Annual EV battery sales reach ~6 TWh (STEP) and ~12 TWh (SD) by 2050, contingent on average pack sizes (~66 kWh BEV; ~12 kWh PHEV) and lifespans.
• 2050 primary demand (STEP) by chemistry scenario relative to 2020: Li increases 17–21× (0.036 Mt to 0.62–0.77 Mt); Co 7–17× (0.035 Mt to 0.25–0.62 Mt); Ni 11–28× (0.13 Mt to 1.5–3.7 Mt). Demand for Co and Ni is substantially lower in LFP and Li–S/Air scenarios due to reduced NCX shares. Al, Cu, graphite are slightly higher in LFP vs NCX (lower specific energy); Li–S/Air uses less Al and Cu per kWh and typically no graphite.
• Cumulative demand 2020–2050 ranges: Li 7.3–18.3 Mt; Co 3.5–16.8 Mt; Ni 18.1–88.9 Mt across fleet and chemistry scenarios. SD about doubles totals vs STEP; NCX roughly 2–2.5× higher Ni/Co cumulative demand vs LFP or Li–S/Air, yielding ~4–5× spread between SD–NCX and STEP–LFP/Li–S/Air.
• EoL batteries and recycling: By 2050, EoL batteries contain ~0.21–0.52 Mt Li, 0.10–0.52 Mt Co, 0.49–2.52 Mt Ni in 9–27 Mt of packs. With hydrometallurgical recycling (NCX/LFP) and mechanical Li recovery (Li–S/Air) and no second-use, recycling can reduce cumulative primary demand by ~20–23% for Li compounds (8% for Li metal), 26–44% for Co, and 22–38% for Ni by 2050. Second-use delays reduce near-term recycling availability.
• Closed-loop recycling potential (CLRP): <10% in 2020s but rises to 20–71% in 2040–2050 depending on material and scenario. CLRP for Li and Ni does not exceed 31% in NCX/LFP due to continuing growth, but exceeds 50% for lithium compounds (and Co up to 71%) in Li–S/Air by 2040–2050; Li metal CLRP remains ≤~10% due to rapid growth and limited historical stock.
• Supply capacity and reserves: Li and Co demand may exceed 2019 production capacities before 2025; by ~2040, EVs could consume an amount comparable to 2019 global primary Ni production. Known reserves of Li, Ni, and Co could be depleted before 2050 in SD (and for Co also in STEP), whereas other materials’ reserves are sufficient to 2050. Graphite and Si production concentration (notably in China) poses additional supply risk; synthetic graphite’s rising share may mitigate some risk.
• Sensitivities: Required battery capacity per vehicle is a dominant uncertainty; extremes (10 kWh all-PHEV vs 110 kWh all-large BEV) create wide demand bounds. Higher specific energies for Li–S/Air could cut cumulative lithium demand by ~20% and Li metal demand by ~40%. High shares of LFP or Li–S/Air, or breakthroughs in post-Li chemistries (e.g., Na, Mg, Ca), could decouple demand from Co/Ni (and potentially Li).

The analysis quantifies how EV uptake, battery chemistry choices, and EoL pathways drive primary demand for critical materials. It shows that under continued dominance of high-Ni NCM/NCA (NCX), Li, Co, and Ni demand rises steeply, pressuring supply chains and known reserves. Shifting toward LFP or Li–S/Air reduces Co and Ni dependence substantially and can improve closed-loop potential mid-century due to slower growth in NCX demand. Recycling, while essential, has limited impact before 2050 because rapid stock growth delays the return flow of EoL material; second-use further postpones availability. Thus, near- to mid-term strategies must focus on scaling responsible primary supply, improving battery specific energy and durability, prioritizing chemistries with abundant elements where feasible (e.g., LFP), and accelerating development of closed-loop processes (especially direct recycling) to capture higher-quality cathode materials. The findings underscore the importance of policy measures (standards, collection, eco-design) and R&D to alleviate supply risks and ensure the EV transition proceeds sustainably.

This study provides global, scenario-based projections for EV battery material demand through 2050, integrating EV fleet growth, evolving chemistries, recycling, and second-use. It highlights substantial increases in Li, Co, and Ni demand in NCX-dominated pathways, the mitigating effect of LFP and Li–S/Air adoption on Co/Ni, and the bounded role of recycling before 2050 due to stock build-up and second-use delays. The results inform supply chain planning, resource governance, and technology roadmaps, emphasizing needs to: (1) expand and diversify sustainable primary supply (especially Li, Co, Ni); (2) improve battery specific energy and lifespans; (3) promote lower-Co/Ni or abundant-element chemistries where suitable; (4) scale effective closed-loop recycling—particularly direct recycling—and enable it via standardization and design-for-recycling; and (5) develop robust collection and second-use frameworks. Future work should refine regional analyses, incorporate heavy-duty and other sectors, assess economic feasibility and purity requirements for recycled materials, and evaluate emerging post-Li chemistries’ system-scale impacts.

• Scope limited to global light-duty EVs; potential additional demand from heavy-duty vehicles and other sectors is not included.
• Significant uncertainties in EV adoption trajectories, average battery capacities, and lifespans; sensitivity analyses show wide demand ranges.
• Battery chemistry market shares (e.g., LFP, Li–S/Air timing and penetration) are speculative; Li–S/Air commercialization, cycle life, and safety remain uncertain.
• Material intensity estimates rely on BatPaC defaults and literature; real-world designs may differ.
• Recycling assumptions include 100% collection and idealized process selection; actual collection rates, economic viability (especially for Li recovery, Al closed-loop, and graphite), and material purities may limit practical recovery.
• Second-use rates and 10-year lifetime are simplifying assumptions; actual repurposing feasibility, performance, and economics vary by chemistry and application.
• Reserve comparisons do not account for future discoveries or changes in resource classification and extraction technologies.