Housing is an important modifier of outdoor environmental hazards due to features such as the amount of passive and active ventilation a dwelling receives, the proportion of the façade that is glazed and the building’s thermal insulation levels. Using Artificial Neural Networks based on an indoor building physics model, we simulate indoor temperature and air pollution concentrations in ~15 million English and Welsh dwellings and assess how exposure to indoor hazards varies for different population groups. The model is derived using simulations from the dynamic thermal modelling tool EnergyPlus, taking a spatially-distributed housing stock as input. Results are linked to the latest 2021 Census data on area-measures of population demographics to assess if vulnerable subgroups bear a disproportionate risk from indoor environmental hazards. We find neighbourhoods in England and Wales with a higher proportion of infants, ethnic minorities and income-deprived populations experience higher two-day maximum indoor temperatures in summertime; whilst more ethnically diverse areas have elevated annual average indoor concentrations of outdoor-sourced PM2.5. Areas with a higher proportion of those aged 65+ had a lower standardised indoor temperature (SIT) in winter, increasing the risk of fuel poverty. We then implement a stock-wide, home energy retrofit, in line with national decarbonisation targets. Results suggest energy-efficient building interventions may exacerbate heat inequalities without the provision of external shading, but improve population exposure to winter indoor temperatures and indoor concentrations of ambient-sourced PM2.5.
Read full abstract