Downstream: End-of-life and recycling performance

Cobalt's end-of-life system improves substantially across the projection period, but recovery never becomes the dominant end-of-life pathway. Chart 26 shows that the collection rate increases by approximately 18% between 2020 and 2050, and recycled volumes overtake downcycled volumes around the 2030s, peaking at almost 80 thousand tonnes. However, uncollected and disposed of material remains the largest end-of-life destination throughout the projection, still exceeding functionally recycled volumes by 2050. This highlights that recycling improvements occur at a rate that does not keep pace with the volume of material reaching end-of-life.

The results in the figure mask variations between applications. Recovery performance depends largely on where cobalt is used rather than on the material itself. By the end of the projection period, cobalt contained in superalloys achieves the highest end-of-life recycling rates of any cobalt-containing application, approaching 80%. This is reinforced by high recycling rates in aerospace and industrial equipment, including gas turbine engines and aircraft, where components are typically large, individually valuable, and handled by specialist processors.⁹⁹ Cobalt contained in permanent magnets is recovered at a much lower rate, while cobalt used in pigments, paints, ceramics, and coatings is effectively unrecoverable because it is chemically dispersed rather than incorporated into recoverable components. Consumer electronics present a different challenge: although personal devices contain relatively small quantities of cobalt, they represent the largest single source of cobalt loss because many never enter formal collection systems.¹⁰⁰ Cobalt's collection challenge is therefore not a single problem but a series of application-specific ones, requiring different recovery approaches rather than a single generic intervention.

The figure below shows that old scrap dominates cobalt's secondary supply throughout the projection period, with its share continuing to increase over time, while new scrap remains comparatively small in volume. This differs from the picture often presented for battery recycling, where manufacturing scrap is currently the most visible secondary source.¹⁰¹ This is not a contradiction—it reflects a difference in what is focused on. The model captures all cobalt applications, including the long-established recycling loops for superalloys and hardmetals, whereas much of the established market evidence focuses specifically on battery materials, where manufacturing scrap continues to dominate because relatively few batteries have yet to reach end of life.¹⁰²

The two scrap streams also differ in how they are recovered. New scrap carries high value, requires no pre-treatment, and can often be returned directly to production, making internal recycling relatively straightforward. Old scrap, by contrast, must first be identified and categorised before it can enter a smelter alongside primary feedstock. For batteries in particular, identifying cell chemistry remains a persistent processing obstacle,¹⁰³ while recovering cobalt from spent lithium-ion batteries requires large, capital-intensive recycling facilities whose economics depend on international trade in spent batteries and black mass.¹⁰⁴

Both the Recycling Input Rate, which stands at 10% in 2020, and the End-of-life Recycling Rate, at approximately 21% (shown in Chart 28), begin the projection from relatively low levels. This reflects the rapid expansion of primary cobalt supply during the battery boom, which dilutes the contribution of secondary material. Both indicators increase substantially over the projection period, with the Recycling Input Rate rising to 34% and the End-of-Life Recycling Rate to 35%. However, part of this improvement reflects declining primary supply rather than exceptional growth in secondary supply. Consequently, a rising Recycling Input Rate driven by a shrinking primary supply base represents a different form of progress than one driven solely by increased recycling.

The model also shows secondary cobalt being generated in greater quantities than the system absorbs during the 2020s, with this gap narrowing only as primary supply tightens in the early 2030s (in the cobalt Supply-demand balance and resilience section). This is consistent with wider evidence that recycled cobalt currently satisfies only a small fraction of global demand and that, while primary cobalt remains readily available, commercial incentives to prioritise secondary material remain limited.¹⁰⁵ Secondary cobalt appears to become economically attractive only when primary supply becomes constrained.

Two structural features of this emerging secondary system deserve particular attention because they reinforce, rather than resolve, the criticality issues identified earlier in this profile. The first is timing. Large volumes of end-of-life electric vehicle batteries only begin entering the recycling system from the late 2020s onwards, meaning cobalt's most significant future secondary resource becomes available only after supply pressures have already emerged. The second is concentration. Battery recycling capacity, together with the trade in black mass that supplies it, also remains concentrated in a small number of countries, with one country accounting for around 80% of global pre-treatment and post-treatment capacity.¹⁰⁶ As a result, secondary supply risks reproducing many of the same geographic dependencies that characterise primary cobalt production. Moreover, secondary cobalt is not automatically free from governance concerns. Informal e-waste collection continues to present documented labour risks, while the complexity of tracing recovered material creates opportunities for primary cobalt to enter supply chains labelled as secondary.¹⁰⁷