Quantifying spatio-temporal carbon intensity within a city using large-scale smart meter data: Unveiling the impact of behind-the-meter generation

This study introduces a novel method for calculating spatio-temporal carbon intensity variations within a city using smart meter data. By integrating smart meter data with solar radiation data from weather satellites, the method predicts electricity demand and solar power generation across 1-km grid areas, achieving higher spatial resolution for carbon intensity distribution than existing models. Accounting for behind-the-meter self-consumption enables dynamic visualisation of carbon intensity variations driven by renewable energy adoption in localised urban areas, offering a more detailed assessment compared to conventional methods focusing solely on temporal fluctuations in the grid's energy mix. The method was applied to a dataset of approximately 410,000 smart meters in Utsunomiya City, Japan. Findings reveal that carbon intensity variations are affected by weather and seasonal changes. Notably, suburban areas with a higher proportion of prosumers exhibit lower carbon intensity than urban centres, highlighting significant intra-city variations linked to local renewable energy utilisation. This method can enhance the efficient use of distributed energy resources within cities and support prioritising low-carbon renewable energy through strategies such as demand response program development, optimising electric vehicle charging schedules, and identifying priority areas for photovoltaic and battery storage deployment.