Lithium-ion battery carbon footprint: understanding hotspots and reducing it

The battery is the most scrutinised component of the energy transition. Electric vehicles, stationary storage, light mobility: it is everywhere. And everywhere, the same question comes up: what is its actual environmental impact? The impact measured, quantified and documented by science, rather than imagined in pub debates ("batteries pollute more than diesel").

The short answer: a battery’s carbon footprint varies by a factor of 2 to 3 depending on where it is manufactured, the chemistry used and the factory’s energy source. Intercontinental transport accounts for only 3 to 5% of the total. To get a sense of scale, see our carbon orders of magnitude. It is counterintuitive, and exactly the kind of result that life cycle assessment is designed to reveal.

This article reviews the real impact hotspots, reduction measures that work, environmental criteria beyond carbon and what the European Batteries Regulation requires. No unnecessary jargon: orders of magnitude from recent scientific literature and our work in the field. For the footprint of an entire car, including its battery, see our article on the environmental impact of electric cars.

1Lithium-ion batteries: what exactly are we talking about?

Before discussing carbon footprints, let us take a moment to recall exactly what we measure. A lithium-ion battery is an assembly of components, rather than a single monolithic object. Each has its own supply chain, geography and share of the environmental impact.

From cell to pack

The basic unit is the cell: two electrodes (a cathode and an anode), an electrolyte and a separator. This is where the electrochemical reaction takes place. Several cells assembled together form a module. Several modules, a management system (the BMS), a casing and a cooling circuit form the battery pack: the finished product incorporated into a vehicle or storage system.

Why does this matter for the carbon footprint? Because the cell accounts for 70 to 80% of the complete pack’s impact. It contains the critical materials (lithium, nickel, cobalt, graphite), and its manufacture consumes the most energy. Reducing a battery’s footprint is essentially about what happens at cell level.

The main chemistry families

What mainly distinguishes one battery from another is its cathode chemistry. Three families dominate the market, and their environmental profiles differ considerably.

• NMC (nickel-manganese-cobalt): the most widespread in electric vehicles and industrial systems. Good energy density (180-250 Wh/kg), but dependent on nickel and cobalt, two materials whose extraction has a heavy impact.
• LFP (lithium-iron-phosphate): growing rapidly, already 80% of the Chinese market. No cobalt or nickel. Lower energy density (120-160 Wh/kg), but longer life (3,000 to 5,000 cycles compared with 1,000 to 2,500 for NMC) and a structurally lower carbon footprint.
• Sodium-ion: the emerging alternative. CATL launched mass production in late 2025. No lithium, cobalt or nickel. Energy density remains modest (175 Wh/kg), but materials are abundant and geographically dispersed. Relevant for stationary storage and light vehicles.

We will return to this in detail in the section on reduction measures, but remember this already: chemistry selection is an environmental decision, as well as a technical or economic one.

2Breaking down a battery’s carbon footprint: the real hotspots

This is the central question of this article, and the answer almost always surprises people. A battery’s carbon footprint is measured in kgCO2e per kWh of energy supplied over its entire lifetime: the unit that allows batteries of different sizes to be compared. The reference scope under the Batteries Regulation is cradle-to-gate, with an optional recycling credit. Breaking down this figure reveals that impact sources are distributed very differently from what we imagine.

Materials: by far the largest source

Extraction and refining of active materials (cathode, anode, electrolyte, separator) account for 40 to 60% of the total footprint. This is the dominant source. For NMC chemistry, nickel and cobalt are the two main contributors: their extraction is energy-intensive and geographically concentrated (DRC for cobalt, Indonesia for nickel), and refining processes have a heavy impact. The graphite anode (synthetic or natural) also contributes significantly.

This is essential to understand: much of a battery’s footprint is determined before the cell even exists, in mines and refining plants. It is a source over which a European assembler has little direct influence, except through chemistry and supplier selection. This is also what a carbon footprint assessment of an industrial site consistently reveals when we examine purchasing closely.

Cell manufacturing: the source we control

Manufacturing itself (dry room, electrochemical formation, ageing) accounts for 15 to 25% of the footprint. It is an extremely energy-intensive process. The dry room runs 24h/24 with massive air-conditioning needs. Electrochemical formation (the first charge-discharge cycles that "activate" the cell) consumes electricity for days.

The consequence is direct: the factory’s electricity source changes everything. The same NMC cell manufactured in Sweden (a low-carbon mix) has a footprint 40 to 60% lower than one manufactured in China (a coal-dominated mix). This is more than a specialist detail: it is the first decarbonisation measure on which a manufacturer can take concrete action.

Transport: the wrongly blamed source

This result consistently surprises our clients. When we support a manufacturer with its first battery LCA, the instinct is always the same: "the cells come from China, so transport must be the problem". In reality, intercontinental sea transport accounts for 3 to 5% of the total footprint. What matters in a battery made in China is the coal powering the manufacturing plant, rather than the Shanghai-Rotterdam journey.

3Key figures: carbon footprint by chemistry and country

Figures circulating about battery carbon footprints are often dated and sometimes contradictory. Here is what the most recent and robust sources say: those on which an industrial decision can be based.

The most comprehensive meta-analysis to date was published in Nature Communications in late 2024. It compiles hundreds of studies and provides median footprints by chemistry: around 74 kgCO₂e/kWh for NMC, 62 for LFP and 82 for NCA. But these medians conceal enormous variation: from 50 to over 110 kgCO₂e/kWh depending on geography and production conditions. The difference mainly comes from the manufacturing electricity mix and the origin of materials.

The JRC (the European Commission’s Joint Research Centre) published a standardised calculation methodology for industrial batteries in 2025. It is the reference for the regulatory declaration under the Batteries Regulation. An important point: this methodology identifies material production and large-scale manufacturing as the two life cycle emission hotspots. The diagnosis is unchanged, but standardisation finally allows like-for-like comparisons.

4How to reduce a battery’s carbon footprint

The good news is that a battery’s footprint is not inevitable: manufacturers have concrete measures that can be ranked and quantified. Here are the four main ones, in descending order of impact.

The manufacturing electricity mix

Of all available measures, this has by far the greatest impact. Studies converge unequivocally: producing cells in a country with a low-carbon grid (Sweden, France, Norway) reduces the footprint by 40 to 60% compared with production in China, using the same technology. This is a fundamental parameter for any European industrial reduction strategy.

European gigafactories are starting to produce at scale. Verkor inaugurated its Dunkirk factory in late 2025, powered by the low-carbon French mix. ACC (Douvrin) is ramping up. These projects concern sovereignty, but they are also decarbonisation measures that can be measured in the LCA. For an assembler, choosing a European cell supplier mechanically improves its carbon score. And with the arrival of the digital product passport in February 2027, this score will be public and comparable.

Chemistry selection

Switching from NMC to LFP chemistry reduces the footprint by around 15% (median 62 vs 74 kgCO₂e/kWh). The difference seems modest, but it is structural: it combines with the geographical measure. An LFP battery manufactured in France combines both advantages. The trade-off is lower energy density: the pack is heavier for the same capacity. For stationary storage, this often does not matter. For vehicles, it is a compromise to assess case by case.

For sodium-ion, the footprint is comparable to LFP, with a materials advantage: the hard-carbon anode emits considerably less than the synthetic graphite used in lithium-ion batteries, and sodium is abundant. An LCA can quantify the trade-off precisely for each application.

Lifetime and recycling

Extending lifetime reduces the footprint per kWh actually delivered over the product’s entire life. A pack that goes from 4,000 to 8,000 cycles divides its normalised footprint by two, without changing manufacturing. This is the effect of the functional unit in LCA, and a measure directly controlled by battery system designers.

Recycling can reduce the total footprint by 5 to 15% through material credits. The Batteries Regulation sets progressive recycled content milestones: a mandatory declaration from 2028, then minimum thresholds (16% recycled cobalt, 6% recycled lithium and nickel) in 2031. For companies subject to BEGES, these data also feed the regulatory declaration. Manufacturers that secure a supply of recycled materials now gain a head start on compliance and their LCA score.

5Beyond CO₂: multicriteria life cycle assessment

The carbon footprint is the indicator that receives the most media coverage, but it is not the only environmental criterion that matters. A battery with a good carbon footprint can have a poor impact on other dimensions, and vice versa. This is the value of multicriteria assessment: it shows the full picture.

Cobalt extraction in the DRC raises human toxicity and ecotoxicity issues that the carbon indicator alone does not capture. Lithium extraction in South American salt flats has a major impact on water resources. Indonesian nickel extracted through laterite processing is associated with deforestation and releases into aquatic environments. These impacts do not disappear because the cell is manufactured in Sweden using low-carbon electricity.

The European PEF (Product Environmental Footprint) methodology, which is gaining ground for batteries, assesses 16 impact indicators: climate change, but also mineral resource depletion, acidification, eutrophication, water use, fine particles and more. This multicriteria approach enables informed trade-offs. Switching from NMC to LFP improves both the carbon and toxicity scores (no cobalt or nickel). But it does not change lithium’s impact on water resources.

6What the Batteries Regulation requires for carbon footprints

All the measures we have just described are moving from good practice into regulatory requirements. Regulation (EU) 2023/1542 is the first legislation in the world to require a verified carbon footprint declaration and a maximum threshold for a manufactured product. It is accompanied by the CSRD, which also requires climate reporting from companies with more than 1,000 employees and €450 million in turnover, multiplying client requests for scope 3 data.

The timetable has slipped. The carbon footprint declaration for EV batteries, compliant with the JRC methodology and verified by a third-party body, was due to apply in February 2025, but will only be required 12 months after the methodological delegated act enters into force. This is still pending: late 2027 at the earliest. Industrial batteries above 2 kWh, originally targeted for February 2026, face the same slippage (18 months after their own delegated act). The digital product passport will be mandatory in February 2027.

But the most consequential deadline comes afterwards. The Regulation provides for a maximum carbon footprint threshold for EV batteries, applicable in February 2028 at the earliest and 18 months after the delegated act setting it. Batteries exceeding it will no longer be allowed onto the European market. The exact figure has not yet been published, but if it falls within the range of current footprints, some Asian manufacturing chains will have to decarbonise or lose market access. Our complete guide to the Batteries Regulation details the full timetable. For a European manufacturer, reliable carbon footprint data and a favourable score become a competitive advantage that can be substantiated, rather than just a marketing argument.

7Starting your battery LCA: where to begin in practice

Do you manufacture or assemble batteries and need to structure your carbon footprint approach? Here are the steps we recommend, in order. Each creates the conditions for the next.

Step 1 - Clarify your category and role

The Regulation distinguishes 5 battery categories with different requirements. The 2 kWh threshold is critical for industrial batteries: below it, the substantial requirements do not apply. And responsibility for the declaration rests with the operator placing the battery on the European market. If you assemble packs from imported cells, you are the declarant, rather than your supplier. This is the most frequent misunderstanding of the Regulation, and probably the most costly if it is not identified in time.

Step 2 - Start collecting supplier data

This is the most time-consuming step. We have supported French assemblers that needed 8 to 12 weeks to obtain data from their cell suppliers: exact chemistry, factory electricity mix and origin of materials. Some suppliers respond quickly with a structured file; others send a table in Mandarin without the origin of the lithium. In our experience, planning this step several months ahead makes all the difference between a calm project and one in panic mode.

Step 3 - Carry out the LCA in accordance with the JRC methodology

Once the data are available, budget €12,000 to €25,000 excl. VAT for a first complete battery LCA (our article on how much an LCA costs details ranges by profile). A second LCA of a similar product costs 30 to 50% less because the model is already in place. Public grants such as Bpifrance’s Diag Éco-conception can fund 60 to 70% of the cost. An important point: this LCA serves more than the carbon footprint declaration. It also feeds the DPP, ESPR eco-design and responses to client requests for scope 3 data. One investment covering four or five requirements.

Step 4 - Manage reduction measures

The LCA identifies the hotspots. The eco-design work begins: negotiating with the supplier on the origin of energy, assessing a chemistry change, incorporating recycled content and extending lifetime. Each change can be quantified again in the LCA model. It is an iterative process, and each score improvement translates into a regulatory and commercial advantage. Start as soon as the data are available!

8Key takeaways

A battery’s carbon footprint is not inevitable: it is a design parameter over which manufacturers have real influence. Materials and manufacturing energy alone account for 80% of the impact, while transport, which attracts imagined concerns, contributes only 3 to 5%. The manufacturing country’s electricity mix and chemistry selection are the first two reduction measures. And multicriteria assessment reminds us that carbon does not tell the whole story: toxicity, water consumption and mineral resource depletion also matter.

The European regulatory framework is turning this knowledge into a concrete requirement. The footprint declaration still awaits the delegated act setting its method, the DPP arrives in February 2027, and the maximum threshold, in 2028 at the earliest, will close the European market to batteries exceeding it. Manufacturers that have invested in LCA and low-carbon sourcing will be on the right side of that line.

The good news is that collection and modelling work serves every purpose: carbon footprint, DPP, ESPR eco-design and client scope 3 data. One investment covers most requirements. So start now, while the preparation window is still open!

Sources: Nature Communications meta-analysis (2024), JRC methodology (2025), Greenpeace East Asia report (2025), Regulation (EU) 2023/1542, Celsius support assignments.

Further resources

Frequently asked questions

1What is the carbon footprint of a lithium-ion battery?

It ranges from 50 to over 110 kgCO₂e per kWh of capacity, depending on chemistry (NMC, LFP, NCA), country of manufacture and origin of materials. The median footprint is around 74 kgCO₂e/kWh for NMC and 62 for LFP (Nature Communications 2024). Our carbon orders of magnitude provide further comparisons.

2Does transport from China really contribute so little to the footprint?

Yes. Intercontinental sea transport accounts for 3 to 5% of the total cradle-to-gate footprint. What matters in a Chinese battery is the coal powering the factory, rather than the Shanghai-Rotterdam journey.

3LFP or NMC: which has the better carbon footprint?

LFP has a footprint around 15% lower than NMC because it contains no cobalt or nickel. But its energy density is lower, making it more suitable for stationary storage than long-distance mobility. The trade-off depends on your application.

4How much does a battery LCA compliant with the European Regulation cost?

Budget €12,000 to €25,000 excl. VAT for a first complete LCA, depending on value chain complexity. A second LCA of a similar product costs 30 to 50% less. Bpifrance’s Diag Éco-conception can fund 60 to 70% of the cost for SMEs.

5Who is responsible for the carbon footprint declaration?

The operator placing the battery on the European market. If you assemble packs from imported cells and sell them in Europe, you are the declarant, rather than your cell supplier. Our Batteries Regulation guide details this point.

6When does the maximum carbon footprint threshold enter into force?

The threshold will be set by a delegated act that the Commission has not yet adopted. It will apply in February 2028 at the earliest, and 18 months after that act enters into force. The exact figure has not yet been published.

7What are the carbon hotspots of a lithium-ion battery?

Active materials are the largest source: their extraction and refining (cathode, anode, electrolyte, separator) account for 40 to 60% of the footprint, led by nickel, cobalt and graphite for NMC. Cell manufacturing (dry room, electrochemical formation) contributes 15 to 25%, and its impact depends mainly on the factory’s electricity mix. Overall, the cell accounts for 70 to 80% of the pack’s impact. Sea transport from Asia contributes only 3 to 5% of the footprint.

8How does the production site’s electricity mix influence the footprint?

It is the dominant parameter: using the same technology, producing cells in a country with low-carbon electricity (Sweden, France, Norway) reduces the footprint by 40 to 60% compared with production in China. Cell manufacturing is very energy-intensive, from the dry room running 24h/24 to electrochemical formation consuming electricity for days. For the same chemistry, the footprint therefore varies by a factor of 2 to 3 depending on the country. Choosing a European cell supplier, such as Verkor in Dunkirk or ACC in Douvrin, mechanically improves an assembler’s carbon score.

9Which measures take priority in reducing a battery’s footprint?

In order of impact, the manufacturing electricity mix comes first: production in Sweden, France or Norway reduces the footprint by 40 to 60% compared with China. Chemistry selection follows: LFP emits around 15% less than NMC (median 62 versus 74 kgCO2e/kWh), a difference that combines with the geographical measure. Extending lifetime also matters: going from 4,000 to 8,000 cycles divides the footprint per kWh delivered by 2. Recycling contributes 5 to 15% through material credits, and minimum recycled content rates will apply from 2031.

Ask Guillaume your questionA personal reply within 24 working hours

Send my questionor: [email protected]