Battery technology and recycling alone will not save the electric mobility transition from future cobalt shortages

The transition to renewable energy and low-carbon technologies requires large amounts of critical materials, including cobalt, which is essential for many lithium-ion batteries used in energy storage and electric vehicles. Cobalt demand has grown rapidly with EV diffusion, while mining and refining are geographically concentrated (e.g., 70% of mine production in DRC and 67% of refining in China in 2019), creating supply risks. Two prominent mitigation strategies are advancing battery technology (reducing or eliminating cobalt) and enhancing recycling. However, their combined effect on relieving global and regional cobalt demand-supply imbalances over time is unclear. This study addresses that question by quantifying historical and prospective global cobalt stocks and flows with regional resolution and by exploring the influence of evolving battery chemistries, lifetimes, and recycling on cobalt demand and supply security.

Prior studies have examined global and national cobalt flows, trade links, demand forecasting, and recycling potential—particularly for lithium-ion batteries. Research has documented cobalt’s growing role in batteries, the uneven global distribution of mining and refining, and the potential for battery development with lower or zero cobalt content. Analyses have also estimated recycling potentials and infrastructure needs. Nonetheless, a integrated global and regional assessment that jointly evaluates how battery technology shifts and recycling progress affect cobalt demand-supply balance across time and regions has been lacking. This work fills that gap by combining dynamic MFA with scenario analysis for battery chemistries, recycling, and primary supply trajectories.

The study applies a dynamic material flow analysis (MFA) of the global anthropogenic cobalt cycle, disaggregated into five regions (China, U.S., Japan, EU, ROW) and five process stages: mining, refining, manufacturing, use, and waste management & recycling. Historical flows and stocks (1998–2019) are quantified using mass balance, process loss coefficients, and lifetime distributions for end-use products. International trade in cobalt-containing final products and old scrap between regions is incorporated. Ten end-use categories are modeled: three emerging (batteries for passenger EVs (B-PEV), electric buses (B-EB), and energy storage systems (B-ESS)) and seven traditional (batteries for consumer electronics & other battery products (B-CE&O), superalloys (SA), cemented carbides (CC), magnets (MAG), catalysts (CAT), pigments (PI), and other uses (OTH)). For prospective analysis (2020–2050), stock-driven MFA models are used. Traditional end-use stocks are projected with logistic per-capita stock trajectories by region combined with population forecasts; inflows/outflows follow lifetime distributions. Emerging end-use demand is computed via a product-specific stock-driven model linking EV and ESS deployment to material intensity: cobalt intensity by battery chemistry, evolving cathode market shares, average battery capacities, EV sales mix (BEV/PHEV), EV market shares, and ESS stock growth. Key parameters and assumptions include: cobalt intensity per cathode type, cathode market share evolution (state-of-the-art NMC/NCA shifting to low-Co NMC-9.5.5/advanced NCA, LFP (zero Co), and later next-gen Co-free (Li-air, Li-S, SSB)), EV market shares, battery capacities, vehicle ownership, ESS growth, battery lifetimes (base and doubled), recycling rates (base: +10% by 2050; high: up to 95%), and population. Four battery technology (BT) pathways are defined: BT1 (state-of-the-art NMC/NCA), BT2 (low-Co NMC-9.5.5 and advanced NCA), BT3 (mature cobalt-free LFP penetrating from 2020 to dominate by 2050), and BT4 (next-gen cobalt-free, e.g., Li-air/Li-S/SSB, penetrating from 2030). Seven combined scenarios (S1–S7) explore mixes of battery tech, lifetime extension, and recycling progress, alongside two primary supply cases: primary-base (announced/scheduled projects to 2030; +1%/yr thereafter) and primary-high (includes unscheduled projects to 2030; +1%/yr thereafter). Regional cobalt supply security is evaluated using domestic reserves and total reserves (domestic plus overseas ownership).

Historical (1998–2019):

• Global cumulative apparent consumption (to manufacturing): 1455 kt; cumulative demand into in-use stocks: 1403 kt.
• 2019 in-use stock reached 471 kt; B-CE&O contributed the largest share (47% of in-use stock; 42% of cumulative demand). Emerging end uses accounted for 3% of cumulative demand and 9% of in-use stock by 2019.
• Primary production supplied 1340 kt (83% of total); old scrap recycled back to refining totaled 279 kt; new scrap recycled to manufacturing was 152 kt (96% from superalloys).
• Cumulative losses: mining 722 kt; refining 162 kt; manufacturing 53 kt; recycling loss 857 kt (with B-CE&O causing 48% of recycling losses due to low collection rates).
• Regional: China had the highest cumulative apparent consumption (541 kt), with 60% in B-CE&O and 216 kt net exports of cobalt in final products. The U.S. and EU consumption emphasized superalloys. For emerging end uses, China dominated cumulative demand and in-use stock (notably 98% of B-EB cumulative demand/stock), with B-PEV the main emerging demand driver.

Prospective (2020–2050):

• Battery technology shifts critically shape cobalt demand. Under S1 (state-of-the-art), cobalt demand for B-PEV reaches 1258 kt in 2050 (79% of total). Under S2 (low-Co), B-PEV reaches 591 kt in 2050 (66% of total). Under cobalt-free scenarios S3 and S4, B-PEV demand peaks at 175 kt (2033) and 612 kt (2038), then declines to 6 kt and 3 kt by 2050 (2% and 1% of totals), respectively.
• Extending battery lifetimes (S5, doubling) reduces cumulative cobalt demand by 619 kt by 2050 (nearly half of S1’s total demand).
• Traditional end uses increase from 144 kt in 2020 to 273 kt in 2050 across scenarios, eventually re-dominating demand as cobalt-free battery technologies penetrate.
• Recycling progress substantially boosts secondary supply: increasing EoL recycling rates to high levels (S6) yields an additional 3680 kt of secondary supply in 2020–2050. Under S4 and S7, secondary supply exceeds total demand after 2044 and 2043, respectively, enabling long-run closure of the cobalt cycle via recycling.
• Primary supply remains essential. Even with best-available tech and recycling (S7), the primary-base supply case results in a supply shortage during 2028–2033. Only with primary-high supply can shortages be avoided under the most technologically optimistic scenarios (S3 and S7).

Regional supply security:

• Domestic reserves in China, U.S., EU, and Japan are collectively <3% of global reserves (7100 kt), implying high risk if relying solely on domestic sources.
• Considering total reserves (domestic plus overseas ownership): China 2383 kt, U.S. 1401 kt, Japan 70 kt, EU 2732 kt, ROW 415 kt. Under these, the EU avoids shortages in S2–S7; the U.S. and China can avoid shortages if cobalt-free techs penetrate (S3, S4, S7). Japan and ROW still face shortages under all scenarios.
• Trade patterns matter: if 2019 trade structures persist, China’s large exports of cobalt-containing products would compromise its supply security even with cobalt-free penetration, whereas the EU secures supply in all scenarios.

Sensitivity analyses show EV market shares and vehicle ownership are key additional drivers of future cobalt demand.

The study directly addresses whether battery technology and recycling can resolve future cobalt supply risks. Results show that while cobalt-free batteries, reduced cobalt intensities, extended lifetimes, and high recycling can dramatically lower demand and increase secondary supply—eventually allowing recycling to meet demand—they cannot avert a short- to medium-term shortage (2028–2033) under plausible primary supply growth. Thus, technology and recycling are necessary but insufficient without increased primary supply. The work highlights the need to accelerate cobalt-free battery R&D and deployment (LFP now; Li-air, Li-S, SSB later), extend battery lifetimes, and rapidly scale recycling systems. Policy and industry implications include: advancing recycling technologies for low-recycled streams (e.g., B-CE&O), implementing extended producer responsibility and design for remanufacturing/reuse/recycling, improving collection and logistics using digital platforms, and establishing regulations and standards to catalyze investment—especially for SMEs. Increasing primary supply through exploration, improved extraction/refining technologies (e.g., HPAL, reopening major mines), and diversified sourcing is critical, acknowledging uncertainties such as price volatility, decade-long project lead times, cobalt’s byproduct dependence on Cu/Ni (94%), declining ore grades, and geopolitical/supply-chain risks. Regionally, China faces high risk due to massive demand growth and limited domestic reserves alongside large refining capacity; diversification of imports, technology shifts to cobalt-free chemistries, and urban mining are vital. The EU and U.S. currently enjoy relatively higher security due to overseas reserve ownership and outsourced battery production but may face changing dynamics as they onshore battery value chains. Japan’s limited reserves and ownership imply higher risk; its hydrogen-focused transport strategy could reduce cobalt dependency.

Advances in battery technology (especially cobalt-free chemistries), longer battery lifetimes, and robust recycling can substantially mitigate long-run cobalt demand and increase secondary supply, potentially closing the cobalt cycle after the 2040s. However, a global cobalt supply shortage between 2028 and 2033 is likely under announced primary supply trajectories, even with optimistic technology and recycling progress. Ensuring the electric mobility transition requires combined actions: accelerate cobalt-free battery R&D and deployment, extend lifetimes, scale high-efficiency recycling and collection systems, and expand/diversify primary cobalt supply with improved extraction/refining efficiency. Future research should refine modeling of additional e-mobility segments (HEVs, two-wheelers, trucks), incorporate battery health and real-time lifetime dynamics, and evaluate circular strategies beyond recycling (remanufacturing and reuse) to further reduce primary cobalt dependence.

The analysis excludes some transport segments (HEVs, two-wheelers, e-bikes, trucks), which may affect demand trajectories. Battery lifetimes are simplified (set as a fraction or equal to vehicle lifetimes) without explicit state-of-health dynamics. End-of-life strategies beyond recycling (e.g., remanufacturing, reuse for grid or private storage) are not explicitly modeled. Data gaps and uncertainties in parameters (technology adoption rates, EV market shares, vehicle ownership, recycling rates, and primary supply schedules) imply that absolute values should be interpreted with caution, though overall conclusions are robust across scenarios.