Recycling rare earths from NiMH batteries looks dirtier on paper. The math explains why.
A new open-access literature review examines greenhouse-gas footprints for producing rare earth oxides (REO) from bastnäsite and monazite versus recovering rare earths from end‑of‑life nickel-metal hydride (NiMH) batteries. When the results are normalized narrowly to rare earth output, the review tallies roughly 85-179 kg CO₂‑eq per kg of recovered REE mixture for NiMH recycling against about 13-22 kg CO₂‑eq per kg of REO for selected primary routes-figures the authors summarize as roughly four‑ to nine‑times higher under simplified assumptions. The comparison is explicitly flagged as not apples‑to‑apples because recycling plants co‑recover nickel and cobalt, use different functional units, and operate under varied energy mixes and system boundaries.
That nuance matters for policy, investment, and procurement. Misreading a multifunctional system through a “REE‑only” lens can make recycling look worse than it is, while obscuring its role in supply security and in displacing fresh mining when all co‑products and avoided burdens are credited. For public buyers and regulators now writing climate and critical-mineral rules into industrial policy, understanding how the math is done is no longer optional background-it determines which projects qualify for green labels and which fall outside the rules.
Where the footprints diverge in practice
- Primary production: Emissions are dominated by concentration, cracking/roasting, and lengthy solvent‑extraction trains to separate REEs-energy‑ and reagent‑intensive steps that scale with ore mineralogy and grade. These are typically located near mines, often on carbon‑intensive grids.
- Recycling flowsheets: Industrial NiMH lines target Ni and Co first, with REEs typically recovered from hydride alloys or residues. Assigning most impacts to REEs can overstate their share when Ni/Co represent larger mass or value and directly displace primary mining of those metals.
- Grid effects: Electricity carbon intensity varies widely across facilities; the same unit process on a coal‑heavy grid will report far higher CO₂‑eq than on a renewables‑rich grid or in plants backed by long‑term renewable power purchase agreements.
- Data granularity: Several recycling LCAs report whole‑plant averages, forcing approximations for REE‑specific burdens and widening uncertainty ranges. For policymakers relying on those LCAs to set thresholds, that lack of resolution can distort which technologies appear “best in class.”
The LCA choices that flip the story
| Modeling choice | Common options | Typical effect on NiMH REE footprint |
|---|---|---|
| Functional unit | kg REO vs. kg mixed REE vs. kg battery processed | Mismatched units can inflate or deflate intensity when normalizing outputs, making cross‑study comparisons-and compliance checks-misleading. |
| Allocation approach | Mass, energy, economic, or system expansion (crediting co‑products/avoided primary) | Economic or system‑expansion methods usually lower the burden attributed to REEs relative to mass allocation, reflecting the fact that nickel and cobalt may carry most of the value and displace upstream mining. |
| System boundary | Cradle‑to‑gate vs. cradle‑to‑grave | Including transport, use‑phase, and end‑of‑life can shift results, especially for avoided mining scenarios and for batteries that travel long distances to centralized recyclers. |
| Electricity and reagents | Regional grid mix; solvent/extractant selection | Cleaner grids and greener separations materially reduce reported GWP and are increasingly shaped by regulation and utility‑scale decarbonization plans. |
| Plant scale and data quality | Industrial metered data vs. lab/pilot estimates | Industrial data with process‑level meters narrows uncertainty, avoids over‑generalization from pilot plants, and provides the kind of auditable trail regulators and investors now expect. |
Policy pressure is converging on harmonized carbon accounting
Battery policy is moving faster than LCA practice. The European Union’s Battery Regulation-now a central pillar of its Green Deal industrial architecture-introduces lifecycle carbon‑footprint rules, a digital battery passport, recycled‑content disclosures and targets, and recovery‑rate benchmarks. Together, these create direct incentives to standardize methods, tighten data governance, and clean up process energy across gigafactories, refiners, and recyclers. The regulation itself, published as Regulation (EU) 2023/1542, is the reference point procurement teams and plant developers must now design against.
- Carbon‑footprint thresholds: For electric‑vehicle batteries, compliance with a maximum lifecycle CO₂‑eq threshold becomes mandatory from 18 February 2028 (or 18 months after the relevant delegated act, whichever is later). Similar obligations phase in for other categories through 2029 and 2031, turning carbon intensity from a reporting metric into an access‑to‑market condition.
- Recycled‑content reporting and targets:
- Documentation of recycled cobalt, lithium, nickel (and lead content in the battery) is required from 18 August 2028 for EV, industrial >2 kWh, and SLI batteries; light‑means‑of‑transport batteries follow in 2033.
- Mandatory minimum recycled‑content shares for cobalt, lead, lithium, and nickel apply from 2031, with higher levels in 2036, effectively underwriting demand for compliant recycling capacity.
- Material recovery targets: By end‑2027 rising to 2031, EU recyclers must meet specific recovery percentages for cobalt, copper, lead, lithium, and nickel-codifying performance baselines that affect process selection, plant design, and ultimately which flowsheets win bankable financing.
- Battery passport: Publicly accessible passport data will include material composition, carbon footprint, recycled content, and due‑diligence information, tightening data integrity across the chain and making inconsistent LCA choices more visible to auditors, watchdogs, and competing suppliers.
What this means for REE supply chains and procurement
For governments, automakers, and utilities signing long‑term offtake deals, the review’s message is that the carbon story of REE recycling cannot be divorced from co‑products and regulatory design. Contracts and industrial‑policy tools need to reflect that complexity rather than rely on a single “kg REE” metric.
- Portfolio decisions should weigh co‑product credits. Where nickel and cobalt displace primary production, system‑expansion or substitution modeling can reveal net‑benefit profiles that a REE‑only snapshot misses. This is particularly relevant for state‑backed investment vehicles and export‑credit agencies tasked with supporting “low‑carbon” supply chains.
- Location matters. Co‑locating hydrometallurgy with low‑carbon power and heat can shift reported GWP materially without fundamental chemistry changes. Site‑selection and permitting agencies now influence carbon intensity as much as process engineers do.
- Specification setting is a lever. Buyers can require LCAs with consistent functional units, declared allocation methods, and process‑level metering-plus third‑party verification aligned to ISO 14040/14044-to ensure comparability. Public tenders and green‑bond frameworks can make such requirements standard rather than optional.
- Technology choices are evolving. Shorter solvent‑extraction trains, greener extractants, membrane separations, and selective precipitation can reduce reagent footprints and wastewater volumes, but only if procurement criteria reward that innovation instead of lowest upfront cost.
Risk and safeguards: where due diligence should focus
The same methodological choices that move academic LCA results also shape real‑world legal exposure-for example, if a plant falls on the wrong side of a future carbon‑footprint threshold or misstates recovery performance in its battery passports. For boards, lenders, and regulators, the risk lens needs to be as granular as the process diagrams.
| Risk area | Why it matters | Practical safeguards |
|---|---|---|
| Allocation transparency | Shifts reported GWP by multiples in multifunctional systems, directly influencing compliance labels and access to green‑finance instruments. | Mandate disclosure of allocation rules; require sensitivity runs with system expansion and economic allocation in both regulatory filings and lender‑mandated LCAs. |
| Data provenance | Plant‑level data vs. generic databases can swing results and undermine trust in disclosed carbon footprints. | Prefer metered industrial datasets; flag proxy fills; audit reagent and energy meters per unit process; embed data‑quality clauses in offtake and financing agreements. |
| Grid and heat intensity | Electricity/steam dominate in separations, and future grid decarbonization pathways may be baked into long‑term contracts. | Source renewable PPAs; electrify heat where feasible; recover and reuse process heat; stress‑test LCAs against alternative grid‑mix scenarios used by regulators. |
| Waste and effluents | Acid/alkali streams and solvent losses drive externalities, permitting risk, and community opposition. | Closed‑loop solvent management; on‑site neutralization; real‑time monitoring; secondary markets for by‑products; transparent reporting to local authorities. |
| Regulatory compliance trajectory | Forthcoming thresholds affect product eligibility and market access, especially under the EU Battery Regulation’s phased‑in rules. | Map plants to EU timelines; align LCAs and passports early to avoid redesign under time pressure; build governance structures that track delegated acts and technical standards as they evolve. |
Interpreting the headline numbers without the traps
Seen in isolation, the review’s REE‑normalized intensities make NiMH recycling look higher‑carbon than primary REE extraction. Read with system context-co‑products, avoided mining, grids in transition, and improved separations-the climate case for recycling can strengthen substantially. The authors underscore that the comparison “is not cleanly comparable because recycling systems are multifunctional,” a reminder that procurement and policy teams need consistent functional units, explicit co‑product credits, and harmonized boundaries before drawing conclusions or writing those numbers into regulation.
How to read and use this review
- Treat the 4-9× figure as a scenario under specific modeling choices-not a universal verdict on recycling. For lawmakers and agencies, it is a caution against hard‑coding a single metric into law.
- Request LCAs that report both REE‑only allocations and whole‑system results with co‑product credits to bound the range, and require that documentation as part of large‑scale subsidy, permitting, or public‑procurement processes.
- Prioritize plants demonstrating cleaner power, greener reagents, and verified recovery rates that map to imminent EU thresholds; for multinational suppliers, that includes aligning disclosure practices with the data fields required by the EU’s battery passport regime.
For now, the review functions as both a technical map and an early compliance stress test: it shows where today’s LCA practices risk under‑ or over‑crediting NiMH recycling just as carbon‑based trade rules, industrial‑policy subsidies, and green‑finance taxonomies are taking shape. Readers across ministries, corporates, and financial institutions will need to read the math as closely as the marketing claims that surround it.
Keep reading
