Comparative Life Cycle Assessment of lithium-ion, sodium-ion, and solid-state batteries for grid-scale energy storage

Grid-scale battery storage is needed to balance the intermittency of solar and wind power. Several battery chemistries compete for this application, but their environmental performance has not been compared within a single, con- sistent framework that covers the full life cycle. This thesis fills that gap by comparing four chemistries — lithium iron phosphate (LFP), nickel man- ganese cobalt (NMC811), sodium-ion (NaNFM442), and pyrite-based solid-state (FeS2/Li/P2S5) — using Life Cycle Assessment with SimaPro, ecoinvent 3.10, and the ReCiPe 2016 midpoint method. The analysis covers three system bound- aries: cradle-to-gate (production), cradle-to-use (including grid charging over the battery lifetime), and cradle-to-grave (including end-of-life recycling). The cradle-to-gate results show large di!erences: NMC811 has the lowest global warming potential (56.8 kgCO2-eq/kWh capacity), followed by LFP (58.5), NaNFM442 (91.3), and pyrite SSB (188.7). However, when the full operational lifetime is considered, the three mature chemistries converge to within 4% of each other (0.376–0.390 kgCO2-eq/kWh delivered). Grid electricity accounts for over 96% of the lifecycle impact. Sensitivity analysis confirms that changing the charging grid from Norway to China causes a 25-fold variation — far larger than any chemistry di!erence. End-of-life recycling changes the results by less than 1%. Beyond climate impact, sodium-ion shows the lowest toxicity and ecotoxic- ity thanks to the absence of copper, cobalt, and lithium. The pyrite SSB’s high production impact is driven by both manufacturing electricity and material syn- thesis; with clean power, its footprint drops by 35%, but remains above that of LFP and NMC due to substantial material contributions. These findings sug- gest that for grid-scale storage, grid decarbonization matters more than battery chemistry selection, and that production-only comparisons can be misleading.