CleanCharge: Emissions-aware electric vehicle charging and infrastructure equity with open data in Melbourne

This study presents CleanCharge, a reproducible, open-data workflow for emissions-aware electric vehicle (EV) charging in Melbourne. The study integrates charger locations with hourly grid carbon intensity obtained using application programming interfaces (APIs) to forecast near-term intensity, identify low-carbon charging windows, and assess spatial equity of access. A streamlined gradient-boosting model delivers high predictive accuracy for hourly carbon intensity (about 94% over a seven-day rolling backtest), enabling practical guidance on when to charge. For a representative three-hour session (20 kWh at 7 kW), shifting into forecasted low-carbon windows cuts per-session emissions by up to 6–7 kg CO2 (around 50% relative to worst-case timing). Clean windows also align with lower costs under both retail and wholesale tariffs. Equity analysis indicates broad metropolitan coverage with persistent access gaps in outer suburbs. All datasets and code are openly available, supporting transparency and reuse. The approach is transferable to other cities and provides evidence for policies that couple charging infrastructure planning with carbon reduction and affordability goals.