EV Battery Pack (75 kWh)

Product Profile

The EV battery pack is the single highest-carbon component in an electric vehicle — and one of the most carbon-intensive manufactured products by total mass-adjusted impact. A 75 kWh pack (the standard capacity in vehicles like the Tesla Model 3 Long Range, BMW i4, and Hyundai IONIQ 6) weighs approximately 450 kg and contains thousands of lithium-ion cells in a structured aluminum or steel enclosure.

At 4,500 kgCO2e per pack (60 kgCO2e/kWh capacity), this represents the central estimate for grid-average manufacturing in the 2023–2025 production window. The IVL range of 61–106 kgCO2e/kWh reflects the enormous sensitivity to manufacturing grid mix: a pack built with 100% renewable energy can reach as low as 40–45 kgCO2e/kWh; a coal-heavy grid pushes it above 100 kgCO2e/kWh.

Why the Score Is What It Is

Battery manufacturing is among the most energy-intensive industrial processes per kilogram of output:

• Cathode synthesis is the dominant hotspot. NMC 811 (nickel-manganese-cobalt) cathode production requires high-temperature calcination at 800–900°C, carbonate precursor co-precipitation, and lithium hydroxide processing. Mining and refining of cobalt (DRC), nickel (Indonesia, Philippines), and lithium (Australia, Chile) each carry significant upstream process emissions.
• Cell manufacturing consumes enormous electrical energy. The formation cycling step — where cells are charged and discharged repeatedly to stabilize the SEI layer — alone accounts for 30–40% of cell manufacturing energy. Dry rooms for electrode handling require constant dehumidification.
• Pack assembly introduces structural materials. An aluminum housing for a 75 kWh pack may weigh 80–100 kg. Primary aluminum at ~8–12 kgCO2e/kg contributes substantially even at modest mass fractions.
• Grid mix dominates variability. CATL’s Ningde facility runs on China’s grid (565 gCO2e/kWh); Tesla’s Gigafactory Nevada uses ~70% renewable energy, reducing cell manufacturing Scope 2 emissions by more than half compared to a China-average baseline.

Scope Breakdown Detail

Scope	kgCO2e	% of Total	Key Drivers
Scope 1	90	2%	On-site calcination furnaces, solvent recovery, coating processes
Scope 2	675	15%	Cell formation cycling, dry room HVAC, electrode coating ovens
Scope 3	3,735	83%	Cathode materials, anode graphite, lithium, cobalt, nickel supply chains
Total	4,500	100%

Chemistry Comparison

LFP (lithium iron phosphate) packs carry a materially lower embodied carbon than NMC packs at equivalent capacity — roughly 20–30% less — because iron phosphate precursors are less energy-intensive to produce than nickel-cobalt-manganese cathode materials, and LFP avoids cobalt entirely. CATL’s LFP cells dominate the Chinese EV market and are increasingly adopted in base-model Western vehicles.

Chemistry	Approx. kgCO2e/kWh	Notes
NMC 811	65–80	High nickel, premium energy density
NMC 622	70–85	Older chemistry, still widely deployed
LFP	45–60	Lower energy density, lower footprint
NCA	65–75	Used by Tesla (legacy), Panasonic cells

Lifecycle Context

Despite the large manufacturing footprint, EV battery packs achieve carbon payback in 1–3 years of driving compared to an equivalent ICE vehicle, depending on the grid mix used for charging. On a European average grid (~257 gCO2e/kWh), a 75 kWh pack’s manufacturing emissions are offset by avoided tailpipe emissions within approximately 18 months of typical driving.

Battery second-life applications (grid storage) and recycling (hydromet recovery of Li, Co, Ni) can further reduce net lifecycle emissions by 15–25%.

Provenance Override

IVL’s 2024 update is the industry reference for battery manufacturing carbon intensity. Tesla’s published Impact Reports provide facility-level data that partially qualify as provenance overrides. Cell manufacturers (CATL, LG Energy Solution, Panasonic) do not publish individual cell PCFs. Battery passport regulations under the EU Battery Regulation (2023/1542) will require verified carbon footprint declarations for EV batteries placed in the EU market from 2027, which will substantially improve data quality.