Towards Net-Zero Computing: Carbon-Aware Energy Scheduling for Data Centres

The rising energy consumption of data centres, driven by artificial intelligence and high-performance computing, poses significant challenges to achieving global net-zero emission targets. This paper proposes a carbon-aware energy scheduling framework that leverages energy generation data to minimize the carbon footprint of data centre operations. Using datasets obtained from ENTSO-E (the European Network of Transmission System Operators for Electricity), the proposed framework dynamically shifts delay-tolerant workloads away from high-carbon periods, effectively aligning computing demand with cleaner energy availability to reduce emissions. Time-series analysis of carbon intensity reveals distinct daily and seasonal patterns that can inform optimal workload scheduling. A correlation analysis between data centre power usage and grid carbon intensity quantifies the potential emission reductions achievable through carbon-aware scheduling. Simulation results demonstrate that the proposed algorithm effectively aligns computing demand with cleaner energy availability, achieving up to 7% carbon reduction under optimal workload flexibility and data centre capacity conditions. These findings highlight the potential of data-driven, carbon-aware scheduling strategies to support sustainable, net-zero aligned data centre operations in the United Kingdom and beyond.