Decarbonizing the lithium-ion battery supply chain

Procurement determines which upstream processes enter a product. Manufacturing determines how inputs become qualified output. Recycling changes the source of materials in the next cycle. Their combined effect must be assessed for the same product.

Identify the main sources of emissions, establish which variables can change, then compare the net benefits of feasible options. Collect data in the same order, prioritizing gaps that could change the decision.

Procurement compares materials that meet performance and supply requirements. Production identifies where energy use and losses occur. R&D assesses whether changes to formulation, design or processing affect life and quality. These questions can share data, but a total footprint cannot answer them all.

Start with a product that has a clear business need. Connect its bill of materials, production line and supply routes so that analysis leads to specific sourcing options, measurement tasks and trials.

Material and manufacturing improvements can first be compared from raw-material extraction to qualified cells at the factory gate. If an option also changes pack design, use-phase efficiency or service life, include those changes in the relevant modules.

Rated capacity, usable capacity and lifetime energy delivered are different denominators. Procurement options for the same specification can be compared at the factory gate. Products delivering the same service require performance, operating conditions and replacement needs to be assessed together. Set the boundary and functional unit before ranking options.

IEA analysis points to different emission profiles for the two chemistries. For NMC, mineral processing and active-material production warrant early attention. For LFP, manufacturing and its energy supply merit close examination.

Shares show where an investigation might begin; they do not rank the total footprints of the two chemistries. Actual sourcing and factory data are needed to establish the hotspots of a specific product.

The same material name does not imply the same production route, energy mix or accounting boundary. Separate these sources of difference before attributing an apparent advantage to production performance.

Qualification, batch consistency and available volume constrain the achievable substitution rate. Only the change that can actually enter the product counts toward procurement-related reductions.

Include copper and aluminum foils, graphite, electrolyte, separators and auxiliary materials alongside the cathode. Preserve the relationship between purchased quantities, process inputs, qualified output and scrap destinations to identify losses and omissions.

Thinner foil or a different formulation may change the processing window, yield and in-use performance. Update the affected manufacturing parameters when assessing an option. Counting only the reduction in one material can miss extra energy or losses; cross-functional trials establish whether those effects change the net benefit.

The China-specific lithium-ion battery dataset compiles 214 processes and 3,197 flows from 53 sources. The associated paper groups the processes into raw-material acquisition and pretreatment, principal product production, and end-of-life treatment and recycling: 75, 52 and 87 processes respectively.

Bills of materials, supply sources and energy measurements should describe actual foreground processes. Match background processes by geography, technology, year and boundary. Existing data provides a traceable starting point for gaps; it does not remove the need for matching or for replacing critical assumptions with measurements.

In the manufacturing model by Degen et al., nickel-based cells require approximately 20-29 kWh of production energy per kWh of cell capacity, compared with 37.5 for LFP. The figures exclude material extraction and refining. They describe a particular model, not a universal factory target or a ranking of total battery footprints.

On site, separate electricity and heat, and identify coating and drying, formation, dry environments and auxiliary loads. Where submetering is absent, start with measurements that explain the main variations rather than expanding reporting fields indiscriminately.

The same quantity of scrap can have different implications before processing and after most manufacturing steps are complete. Record inputs, transfers, rework and rejects at critical stages rather than using a factory-wide yield to explain every loss.

Include avoided losses, rework energy and scrap treatment in the assessment. If higher yield requires longer processing or extra treatment, recalculate the net change per unit of qualified output. An improvement in one percentage alone does not establish the environmental benefit.

Separate energy needed to maintain the environment and auxiliary systems from energy that varies with processing, then relate both to qualified output. Neither category is perfectly fixed or linear in practice: the distinction helps identify relationships that need measurement and testing.

If qualified output grows faster than base energy demand, the burden per unit can fall without an improvement in equipment efficiency. Align product mix, operating hours, environmental requirements and yield to distinguish volume effects from process changes.

Reducing electricity use and lowering its applicable emission factor both affect electricity-related emissions. Adding benefits calculated separately against the original baseline double-counts part of the reduction. Multiply the updated quantity by the updated factor to obtain the combined burden.

Keep physical supply, contractual evidence, applicable periods and accounting rules connected. Corporate purchased-electricity reporting and product footprints can follow different requirements. Buying an electricity instrument does not automatically determine its treatment in every product declaration.

Machala et al. compare industrial recycling and primary-material supply chains. Their gate-to-gate refining results are 14.5, 6.1 and 2.8 kg CO2e per kg NCA-equivalent for the primary route, end-of-life battery recycling and production-scrap recycling respectively. Data is mainly from 2021. Outputs are equivalent in elemental content, but differ in product form.

The findings support separate assessment of feedstock and processing conditions. Production scrap and end-of-life batteries are not interchangeable inputs. Match collection, pretreatment and refining steps to the proposed route, then assess whether its actual conditions support a comparable direction of benefit.

Define the quality, destination and feasible substitution share of each recovered material. Battery-grade use and other applications imply different comparison objects. Recovered output is not automatically equal to avoided virgin input.

Net effects depend on added collection, treatment and transport, together with the burden of the route actually displaced. Align boundaries, time periods and product functions. For multiple recovered products, explain allocation so that the same benefit is not claimed repeatedly by different parties.

When a material or product option affects lifetime, efficiency or replacement frequency, extend the comparison to the same service. Stationary storage requires actual cycling, degradation and operating conditions; traction batteries need to match the vehicle service being considered.

Manufacturing and use can share basic data while retaining distinct boundaries and assumptions. Avoid applying a single per-kWh ranking across applications. The useful result states the uses and operating conditions under which an advantage persists.

Lower-carbon materials may alter processing conditions. Higher yield may reduce material demand. Changes in line operation may affect energy allocated per unit. Update connected parameters rather than changing only the most favorable input for each measure.

Combination analysis reveals reinforcing effects, double counting and added burdens. Record what changed, why it changed and which inputs were held constant so that the model can be matched to a real trial.

Vary supply factors, energy use, yield and substitution shares within plausible ranges and observe whether the ranking changes. A parameter that reverses the conclusion after a small change deserves early verification. Ranges must reflect feasible conditions, not an arbitrary combination of favorable assumptions.

If the ranking remains stable across plausible conditions, proceed to feasibility testing. If it reverses readily, obtain stronger evidence first. This sequence directs data effort toward action rather than assigning equal priority to every gap.

For the same product and production baseline, identify additional investment, energy and material savings, maintenance, qualification and operating changes. Cost and environmental scenarios should share physical inputs while retaining their own price, lifetime and accounting assumptions.

Screen technical feasibility, performance and supply conditions before comparing net reductions and incremental costs. Near-term operating improvements and long-term capital projects need not compete in one ranking. Organize actions by readiness and timescale to connect them to investment and production plans.

The first product should have a clear sourcing or production question, reasonably complete records and accountable owners. Define its specification, factory, time period and decision purpose at the start to prevent scope from expanding during data collection.

Progress depends on evidence and validation conditions. Each stage leaves a transferable output: scope, baseline and gaps, candidate options, and trial results. Schedule qualification, equipment adjustments and lifetime testing around their actual requirements.

Retain original quantities, units, sources, periods, applicable products and owners. Manage calculation relationships and background factors separately, with results tied to input versions. Updates can then distinguish production improvements, data replacement and methodological changes.

Define update triggers for supply sources, production processes and background factors. After a supplier change, a shift in output mix or a data revision, rerun the model consistently and attribute the difference to specific causes. The same data package can then support the next sourcing decision, product iteration or factory improvement.

Use purchasing, inventory, output and waste records to reconcile major material flows, and energy and production records to explain loads. Investigate anomalies before comparing options so that missing data or changing definitions are not mistaken for improvements.

Model checks cover units, duplicated processes, allocation and parameter relationships. External claims need their boundaries, evidence and intended use checked. Confirm the validation requirements for customer communication, comparative claims and regulatory submissions separately.

Start with the product, the current choice and the available data. Are supplier differences difficult to explain? Is manufacturing energy hard to locate? Is the net benefit of a circular route unclear? Define the decision before selecting the processes and evidence to connect.

China-specific unit-process data and product-footprint research provide a published methodological foundation. Connecting that foundation to actual company records can support a baseline, contribution analysis, conditional options and a validation path. Explore the public research and team at Tiangong Discover.

Production comparisons center on qualified cells at the factory gate. Add separate modules for changes to packs, use or end-of-life treatment. For options with different performance or life, establish equivalent service requirements. Match background processes by technology, geography, time and boundary, retaining replacement records.

Cite literature results with their original boundaries and data periods; recalculate company options using actual product and process data. Retain scenario inputs, equations and conditions, distinguishing production changes from data updates. Stage shares describe contribution patterns, not the ranking of total product footprints.

The start of the EU battery passport and the obligations for different information fields have distinct applicability dates. Check product categories, carbon footprints, recycled content and other fields separately instead of assigning one date to every obligation.

Manage each field through its source, owner, update frequency and access permissions. Once the reporting purpose is defined, verify the applicable calculation method, validation procedure and legal text, and maintain version links to the internal improvement model.

Confirm the product and scope, then select a variable that can change. A material substitution, process improvement or circular route can be the starting point. Preserve a chain of evidence from inputs through changes to results.

If an advantage depends on an unverified condition, measure or test it next. If it remains stable within plausible ranges and meets quality and supply requirements, proceed to the corresponding implementation plan.