Are EVs worse for the climate once battery production is included?

Battery production gives a battery-electric car a larger manufacturing footprint than a comparable combustion car in many lifecycle assessments, and that is a real climate cost. It does not, by itself, determine which car has the lower emissions over its whole life.

Claim review

Claim: “EVs are worse for the climate once battery production is included.”
Verdict: Mostly incorrect
Scope: Current passenger battery-electric vehicles compared like for like with new combustion cars. The global conclusion is illustrated with IEA analysis and current European and U.S. studies; the size of the benefit remains regional and model-dependent.
Short answer: Full lifecycle studies do include battery production, and they generally find higher manufacturing emissions for a BEV. For comparable passenger cars driven through a normal service life, lower use-phase emissions usually more than repay that initial difference. The margin can be small on carbon-intensive electricity or with a large battery, inefficient vehicle, low lifetime mileage or premature scrappage, so no single percentage or break-even distance applies everywhere.
Last reviewed: 6 August 2026
Review trigger: A new major regional lifecycle assessment, a material update to the IEA calculator or lifecycle model, or evidence that changes battery-production emissions, grid trajectories or assumed vehicle lifetimes.

This article is part of EV Claims, Checked.

What the claim gets right

An EV does not begin life with zero emissions. Mining and processing raw materials, producing active materials and cells, assembling the battery pack, manufacturing the rest of the vehicle and transporting components all require energy. The battery usually makes a BEV more emissions-intensive to manufacture than a comparable combustion vehicle.

The ICCT's 2025 European assessment estimated about 40% higher production emissions for its representative battery-electric car than for its gasoline equivalent, primarily because of the battery. ICCT — Life-Cycle GHG Emissions from Passenger Cars in the EU, 2025 The exact difference is not a property of every EV. It changes with battery capacity and chemistry, factory energy, material source, production yield, vehicle design and the combustion vehicle used for comparison.

The claim is also right that charging is not automatically carbon-free. A BEV has no exhaust CO₂, but electricity generation can produce emissions. Upstream fuel production also matters for a combustion car: extracting, transporting and refining oil is part of a fair comparison, not only what exits the tailpipe.

What a lifecycle comparison must include

“Lifecycle emissions” can describe different boundaries. A credible like-for-like passenger-car comparison should identify at least:

production of the vehicle, battery and major materials;
production and delivery of electricity or liquid fuel;
energy used while driving, based on a stated test or real-world adjustment;
maintenance and replacement assumptions;
vehicle lifetime and total distance driven; and
end-of-life treatment, including how recycling burdens or credits are allocated.

The result is normally expressed as total tonnes of CO₂-equivalent over the vehicle life or as grams of CO₂-equivalent per kilometre. The second form divides a largely upfront manufacturing footprint by an assumed lifetime distance. A car driven 80,000 km and one driven 240,000 km can therefore have very different per-kilometre production results even if they left the same factory.

The U.S. Department of Energy's R&D GREET model includes vehicle and battery production, fuel production and use, facility construction and end-of-life. It also explains why the functional unit matters: the service being compared is transportation over a defined distance, not simply the existence of two vehicles. U.S. Department of Energy — R&D GREET Life Cycle Assessment Model

Battery-only footprints answer a narrower question than whole-vehicle LCAs. The EU Batteries Regulation, for example, defines a battery carbon-footprint boundary that covers raw-material acquisition, battery production, distribution, and end-of-life and recycling. It uses the energy delivered over the battery's expected service life as its functional unit. European Union — Regulation (EU) 2023/1542 Concerning Batteries and Waste Batteries That is valuable battery information, but it cannot alone determine the lifetime result for the complete car.

What current evidence shows

The IEA's global analysis for a medium-size car sold in 2023 includes vehicle production and well-to-wheel emissions over 15 years and about 200,000 km. In its Stated Policies Scenario, the BEV produces about half the lifecycle emissions of an equivalent combustion car. It remains about 30% lower even when the calculation removes future grid improvement. IEA — Global EV Outlook 2024: Lifecycle Impacts of Electric Cars

That global average hides large regional differences. In the same analysis, a medium-size BEV's lifecycle emissions were about 20% below its combustion equivalent in India, where electricity was more carbon-intensive, around 40% lower in China, and about 65% lower in the United States. These are modelled regional results under common IEA boundaries, not labels that can be attached to every vehicle sold in those countries.

Two newer regional assessments show the same general direction under their own assumptions:

The ICCT estimated 63 g CO₂e/km for a representative medium-segment BEV sold in the European Union in 2025 and operated on the projected 2025–2044 EU electricity mix. Its comparable gasoline car was estimated at 235 g CO₂e/km, making the BEV result 73% lower. The extra production emissions were offset after about 17,000 km in that study. ICCT — Life-Cycle GHG Emissions from Passenger Cars in the EU, 2025
The U.S. Department of Energy's representative 2025 electric and gasoline light-duty SUVs, modelled with R&D GREET 2024 and an average U.S. grid trajectory, gave the EV 46% lower lifecycle greenhouse-gas emissions. U.S. Department of Energy — R&D GREET Life Cycle Assessment Model

The percentages are not interchangeable. The studies use different regions, years, cars, electricity trajectories, lifetimes and model boundaries. Agreement about the direction does not create a universal 46%, 50% or 73% saving.

The IEA provides an interactive lifecycle calculator precisely because changing vehicle size, battery size, efficiency, lifetime, annual distance and electricity assumptions changes the result. IEA — EV Life Cycle Assessment Calculator

Why battery production estimates differ

A battery's capacity is an important driver because more cells and materials generally mean more production emissions. Comparing a small-battery hatchback with a large gasoline SUV can flatter the EV; comparing a large long-range electric SUV with a small hybrid can do the opposite. The defensible comparison is between vehicles that provide a similar function, while still showing the effect of choosing a larger vehicle in either powertrain.

Chemistry and supply chain also matter. The IEA estimated pack-level emissions per kWh for lithium iron phosphate batteries to be about one-third lower than for high-nickel NMC batteries in its 2024 analysis. It found different hotspots: critical-mineral processing dominated more of the NMC footprint, while cell and pack manufacturing represented a larger share for LFP. IEA — Global EV Outlook 2024: Lifecycle Impacts of Electric Cars

Factory electricity is only part of the picture. Ore grade, refining route, process heat, material efficiency, production scrap, energy density and transport can all change the result. A single “kg CO₂e per kWh” value copied from one study should not be applied to every battery year, chemistry and factory.

This is also why the production footprint can improve before the vehicle is driven. Cleaner industrial energy, higher material and energy efficiency, smaller packs, less production scrap and recycled inputs can reduce it. None of those improvements makes today's battery-production emissions disappear; they determine the size of the initial emissions debt.

Electricity mix and real energy use

During use, a BEV's climate intensity is approximately its electricity consumption from the grid multiplied by the emissions intensity of the electricity, with the study's treatment of charging losses and infrastructure added where applicable. Both parts vary.

A more efficient EV needs less electricity for the same distance. Cold weather, high speed, towing, large wheels and cabin heating can increase consumption. Charging losses mean electricity drawn from the grid is higher than energy stored in the battery. Any calculation must say which measurement boundary it uses.

Electricity emissions vary by country, region, hour and year. Most attributional vehicle LCAs use an average mix and may model how it changes over the vehicle's life. A marginal-emissions analysis asks a different question: which generators respond to an additional charging load at a given time? Both can be useful, but a figure from one method should not be presented as if it answered the other.

Future grid decarbonisation is an assumption, not a guarantee. It is reasonable for a lifetime study to model the electricity a car is expected to use in later years, but the result should also show what happens if that trajectory is slower. The IEA's finding that its global-average BEV remained lower without future grid improvement is therefore an important sensitivity result, not proof that the grid never matters. IEA — Global EV Outlook 2024: Lifecycle Impacts of Electric Cars

Lifetime mileage and the emissions payback point

The manufacturing difference happens before delivery; use-phase emissions accumulate with every kilometre. An emissions “payback” point is the distance at which the BEV's lower cumulative use emissions have offset its higher production emissions.

That distance cannot be universal. It becomes longer when:

the BEV has a large or emissions-intensive battery;
charging electricity has high emissions;
the BEV consumes more electricity than assumed;
the combustion comparator is unusually efficient;
annual driving is low; or
the vehicle is retired early.

It becomes shorter with a smaller and cleaner-produced battery, lower-carbon electricity, better energy efficiency, higher annual use or a less efficient combustion comparator. The ICCT's 17,000 km European result is useful only with that study's vehicles, grid path and boundary. It is not an expiry date for the debate in every market.

Low mileage does not necessarily reverse the result, but it gives the fixed manufacturing footprint fewer kilometres over which to spread. It also raises a separate purchasing question: replacing a functioning car with any newly manufactured vehicle is not the same comparison as choosing between a new BEV and a new combustion car. A replacement decision should account for the remaining life of the existing car rather than treating its manufacturing emissions as if they had not already occurred.

Battery replacement assumptions

A lifecycle model must say whether the original battery lasts for the vehicle's modelled life. Quietly adding a full replacement pack to the BEV while assuming no major replacement in the combustion car is an asymmetric comparison. Quietly assuming no replacement when a pack actually fails is also too favourable.

Capacity degradation is not the same as battery failure. A battery with reduced usable capacity may continue to power the car, and some faults can be repaired below full-pack level. EVKX examines observed degradation, failure, warranties and repair boundaries in How long do EV batteries really last? and explains the underlying pack in EV Batteries.

Current fleet age limits how confidently very long-term replacement rates can be projected. The IEA reported in 2026 that nearly all EV and stationary-storage batteries deployed during the recent growth period were still in use and that most were expected to operate until the mid-2030s or potentially longer. It described a roughly 15-year lag before comparable volumes reach end of life. IEA — Global EV Outlook 2026: Electric Vehicle Batteries That supports neither “batteries never need replacement” nor an assumption that every EV requires one full new pack.

Sensitivity cases are the honest solution. A study can show no replacement, a partial repair, and one replacement at a stated age, then let readers see whether the verdict changes.

End-of-life and recycling

End-of-life treatment is real but comparatively distant for most EVs now on the road. A lifecycle study should state whether it includes collection, dismantling, transport, recycling energy and credits for recovered material. Different allocation rules can shift emissions between the first battery and the product that uses the recovered material later.

Recycling does not erase the emissions of the original mining and manufacturing. It can avoid part of the future need for virgin material, and recycled inputs can have a lower production footprint depending on the process and electricity. The effect should be counted once, under a declared method.

The IEA expects end-of-life battery volumes to grow materially after 2035; until then, manufacturing scrap is a major source of recycling feedstock. IEA — Global EV Outlook 2026: Electric Vehicle Batteries EVKX reviews the distinction between technical recoverability, actual collection and market capacity in Can EV batteries be recycled? and covers repair, reuse and second-life routes in EV Battery Lifecycle, Repair, Second Life and Recycling.

Climate is not every environmental impact

The claim tested here is about climate, so the main metric is greenhouse-gas emissions expressed as CO₂-equivalent. Mining can also affect water, land, biodiversity, workers and local communities. Combustion fuel supply has its own extraction, refining, spill, methane, air-pollution and geopolitical impacts.

Those concerns should be assessed with suitable indicators rather than hidden inside one carbon number. A lower lifecycle climate footprint does not mean an EV has no environmental cost, and evidence of a mining impact does not by itself prove a higher climate footprint.

What this means in practice

For a buyer comparing new cars, the most defensible climate comparison is between vehicles of similar size and function, using realistic energy consumption, the expected charging region, a plausible lifetime and transparent battery assumptions. Smaller, more efficient vehicles reduce emissions whichever powertrain they use.

For a published claim, ask five questions:

Does it include both battery production and fuel production?
Are the vehicles genuinely comparable in size and use?
Which electricity mix, driving distance and lifetime are assumed?
Is a battery replacement added, and is that assumption supported?
How are recycling and end-of-life treated?

The evidence available in August 2026 does not support the general statement that adding battery production makes EVs worse for the climate. It supports a more precise conclusion: BEVs normally begin with higher production emissions, then recover that difference through lower use-phase emissions over a normal life. How large and how fast that benefit becomes depends on the car, battery, electricity system, mileage and lifecycle method.