Is an EV just a coal-powered car?

Per-km climate impact: charging an EV in each state vs petrol

Each bar is grams of CO₂-equivalent per km for one case: an EV charged from that state’s 2024 grid, using the efficiency and charging-loss inputs below. Shorter is better (less climate impact on this measure). The orange line and shaded zone use your petrol inputs (L/100km and well-to-wheel) — the line is a petrol car, tailpipe plus fuel upstream. Anything to the right of the line is worse per km than that petrol; the coloured EV rows sit to the left in every state at the defaults, meaning lower g/km than this petrol benchmark across Australia.

Your EV

Pack-to-wheel consumption plus a charging-loss allowance gives kWh drawn from the grid per km.

EV consumption (kWh/100km) Small EV 14-16, mid 17-19, large SUV 20-24. Charging loss factor Home AC charging: roughly 1.08-1.12x pack kWh. DC fast charging: closer to 1.05x.

Your petrol reference

Pick a petrol car to compare against. Well-to-wheel = tailpipe + refining/distribution upstream.

Fuel consumption (L/100km) ABS on-road average for the AU fleet sits around 9-10 L/100km. Well-to-wheel multiplier Tailpipe CO2 times this factor approximates refining + distribution upstream. Most AU/IEA figures use 1.15-1.25.

State grids, rounded

State	Grid (gCO2e/kWh)	EV (gCO2e/km)	vs petrol	Mix note

Intensities are rounded annual 2024 averages from OpenNEM / AEMO CDEII. Instantaneous values swing more than the annual headline; a Victorian EV charged at noon under solar sits closer to SA's number than to VIC's coal average.

Methodology

EV gCO2e/km = grid_intensity (gCO2e/kWh) * consumption (kWh/100km) / 100 * charging_loss_factor. A 18 kWh/100km EV on the NSW grid at 580 gCO2e/kWh and a 1.10 loss factor lands at about 115 gCO2e/km.
Petrol well-to-wheel = fuel_consumption (L/100km) * petrol_factor (2.31 kgCO2/L) * 10 * wtw_multiplier. A 9 L/100km petrol car at 1.20 well-to-wheel lands around 250 gCO2e/km; that is the red band on the chart.
Annual vs marginal. The grid numbers are annual averages. The marginal generator at 18:00 on a windless winter evening is usually gas or coal, so the "right now" number is often higher than the annual one. Equally, midday on a sunny day the marginal supply is rooftop PV at roughly zero operating emissions, so the annual average hides a bimodal distribution.
Scope. Tailpipe + upstream for the petrol side; grid + charging losses for the EV side. Neither includes embodied manufacturing - that is what /lifecycle models. Tyre wear and other non-exhaust PM are outside this page; those land in the planned /non-exhaust page.
WA context. The SWIS grid is not part of the NEM; its intensity figure comes from separate reporting (IMO / AEMO WEM). Treat it as comparable in order of magnitude, not pooled into the NEM average.