Modeling the life cycle of an automotive NCM622 battery pack
A literature-based study using HiQLCD and ecoinvent in openLCA to examine a battery pack manufactured in China and used in Europe.
One battery pack. A lifetime of delivered energy.
This study combines published cell and pack inventories with HiQLCD’s China-specific datasets to model a 72.5 kWh automotive NCM622 traction battery manufactured in China and used in Europe. It examines environmental hotspots and background-data effects.
The scenario brings together several literature sources. It does not represent one observed industrial battery design or a measured customer product.
Cumulatively delivered over the battery’s automotive service life.
- Material supplyCathode, anode, collectors & pack components
- ManufacturingCell production & pack assembly in China
- DistributionPackaging & transport to Europe
- UseCharging and discharging losses
- End of lifeCollection, disassembly & treatment in Europe
Use-phase impacts include round-trip electricity losses only. Useful traction electricity, the vehicle and charging infrastructure are excluded. The 72.5 kWh initial capacity is not the denominator of the reported results.
Connecting a literature inventory to background processes
Literature & scenario
Materials, quantities, manufacturing energy and lifetime assumptions.
Selected datasets
HiQLCD and ecoinvent records for the inputs and related activities.
openLCA model
Model connections, EF 3.1 calculation and sensitivity analysis.
| Model input | Background-data choice to inspect |
|---|---|
| NCM622 cathode | Material composition and production route |
| Synthetic graphite | Graphitization process and technology mix |
| Copper & aluminium foils | Material specification and processing boundary |
| Electricity | Production or use location and voltage level |
These are examples of input-to-provider relationships. Model A uses a mixture of backgrounds; its inputs should not all be attributed to HiQLCD.
Materials and electricity shape the climate result
Model A reports 0.3001 kg CO₂-eq per lifetime-delivered kWh. The largest background-provider contributions identify where data selection deserves close attention.
Selected provider-level contributions, not lifecycle-stage shares; the displayed contributors do not sum to 100%.
The same foreground, two background configurations
Model A combines selected HiQLCD and ecoinvent providers. Model B uses ecoinvent backgrounds. Foreground quantities are held constant, while the substitutions also change technology, geography and dataset boundaries.
Net difference, B minus A, in kg CO₂-eq / FU. This compares modeled configurations; it is not an emissions reduction caused by purchasing data.
Why does the result change?
The synthetic-graphite provider change contributes +0.10871 kg CO₂-eq / FU. All remaining substitutions together contribute −0.01151, partly offsetting that increase. The graphite-route difference is larger than the net change.
Compare all 16 EF impact categories
FU is one lifetime-delivered kWh. Fourteen categories increase and two decrease in Model B; climate change alone does not establish overall environmental superiority.
| Impact category | Model A | Model B | B/A − 1 | Unit / FU |
|---|---|---|---|---|
| Climate change | 0.3001 | 0.3973 | +32.4% | kg CO2-eq |
| Ozone depletion | 4.025×10-9 | 4.145×10-8 | +929.9% | kg CFC-11-Eq |
| Particulate matter | 1.868×10-8 | 3.369×10-8 | +80.4% | disease incidence |
| Ionising radiation | 0.03605 | 0.04511 | +25.1% | kBq U235-Eq |
| Photochemical ozone formation | 1.059×10-3 | 1.581×10-3 | +49.2% | kg NMVOC-Eq |
| Acidification | 2.166×10-3 | 3.595×10-3 | +65.9% | mol H+-Eq |
| Terrestrial eutrophication | 3.368×10-3 | 5.320×10-3 | +58.0% | mol N-Eq |
| Freshwater eutrophication | 1.313×10-4 | 1.757×10-4 | +33.8% | kg P-Eq |
| Marine eutrophication | 3.227×10-4 | 5.129×10-4 | +59.0% | kg N-Eq |
| Freshwater ecotoxicity | 3.178 | 3.666 | +15.4% | CTUe |
| Human toxicity, cancer | 1.528×10-8 | 2.627×10-10 | -98.3% | CTUh |
| Human toxicity, non-cancer | 6.119×10-9 | 1.230×10-8 | +101.0% | CTUh |
| Land use | 1.316 | 1.451 | +10.3% | dimensionless |
| Water use | 0.5161 | 0.5596 | +8.4% | m3 world eq. deprived |
| Resource use, fossils | 4.294 | 5.346 | +24.5% | MJ, net calorific value |
| Resource use, minerals and metals | 2.702×10-5 | 2.450×10-5 | -9.3% | kg Sb-Eq |
Read the findings within the study’s boundaries
The main limitations are a composite pack architecture, supplementary-table transcriptions not independently verified, simplified component production and yields, generic treatment routes and missing supplier-specific data.
The study is useful for examining background choices and hotspots. Applying it to another product requires rebuilding the relevant quantities, processes, geography and lifetime assumptions.
Study environment & uncertainty
- Software & method
- openLCA 2.6.2 · EF 3.1 midpoint
- Model A backgrounds
- Lithium-battery industry pack 1.1, HiQLCD 1.5.0 and ecoinvent 3.12 cut-off
- Model B backgrounds
- ecoinvent 3.12 cut-off
- Functional unit
- 1 kWh cumulatively delivered over automotive service life
The 500-run Monte Carlo analysis varies selected foreground distributions in Model A, with backgrounds fixed. Its P2.5–P97.5 range is 0.2361–0.4220 kg CO₂-eq / FU. It does not validate the significance of the A/B comparison or establish a measured product’s confidence interval.
