Manufacturing (embodied) vs 150,000 km lifetime
Bars scale to the larger of the two inputs. Petrol excludes fuel manufacturing here; that enters via well-to-wheel litres/km below.
Cumulative CO2e at 150,000 km (same inputs): — t petrol vs — t EV.
Cumulative CO2e vs distance
Vertical line marks where the EV curve crosses under the petrol curve (when it exists). If your grid is dirty enough that EV g/km exceeds petrol g/km, there is no crossover — that is a real mathematical outcome, not a bug.
| Odometer | EV cumulative (t) | Petrol cumulative (t) |
|---|
Embodied manufacturing
Tonnes CO2e for vehicle production. Battery dominates the EV side; ICE still has steel, aluminium, engine line.
Use-phase
Same defaults as the cost-per-km page for consumption. Grid intensity is the knob people argue about.
Grid intensity, not a second fuel type
A charger does not have its own smokestack. An EV draws power from the same grid as your heat pump and your neighbour’s lights — unless your wall kWh come from rooftop solar, in which case the relevant operational intensity for those kWh is ~0 g CO2e/kWh (manufacturing the panels is a separate LCA line item; we do not model it here). Whether the rest is coal, gas, wind, or grid-purchased renewables is an emissions intensity question — grams CO2e per kWh — not a yes/no “fossil yes fossil no” switch. Public DC vs home AC changes who bills you and losses; we still treat DC energy as drawing the grid-average slider unless you split intensities yourself. For policy work, modellers argue about average vs marginal grid intensity; this page uses a blunt average knob plus an explicit solar-at-home share so you can see how much the answer moves when you charge off your own array.
Methodology & sources
What this model does
eff_g_per_kWh = dc_share × grid + (1 − dc_share) × ((1 − solar_home) × grid + solar_home × 0)
EV_g_per_km = (kWh/100km) × (1 + charger_loss) × eff_g_per_kWh
ICE_g_per_km = (L/100km) × WTW_kg_CO2e_per_L × 1000
cumulative_t(vehicle, km) = embodied_t + (g_per_km × km) / 1e6
Crossover km solves embodied_ev + ev_g·k = embodied_ice + ice_g·k for k when ev_g < ice_g and embodied_ev > embodied_ice. If embodied_ev ≤ embodied_ice already, the EV wins from km 0. If ev_g ≥ ice_g, there is no finite crossover.
Peer-reviewed / agency references (defaults are rounded)
- ICCT 2021 — global passenger car lifecycle GHG. Synthesis across regions: for most grids, BEV lifecycle emissions fall below comparable ICE over vehicle life; manufacturing gap is repaid in the tens of thousands of km, not “never”. Exact breakeven is sensitive to class, battery size, and grid year.
- DCCEEW National Greenhouse Accounts Factors. Official Australian emission factors for fuels and electricity (updated periodically). Use them if you need report-grade numbers; the slider here is a coarse teaching default.
- Petrol combustion factor in Australian accounts is about 2.31 kg CO2-e per litre combusted (scope 1 from energy content). Well-to-wheel adds upstream extraction and refining; we default 2.65 kg/L as a mid estimate — edit the field if you prefer combustion-only.
What this page deliberately does NOT do
- Regional marginal intensity hour-by-hour (would need dispatch modelling).
- Maintenance, tyres, road wear, or brake particulates (real, but a different page).
- Second-life battery credit or hydrofluoric acid from recycling (could lower net EV manufacturing over long horizons).
- Hybrid or PHEV pathways.
- Embodied emissions from installing rooftop PV (only operational kWh are free of CO2 here).
- A different grid intensity for public DC than for home imports (both use the same slider today).