The COVID-19 pandemic has affected us all significantly, leading notably to school closures for lengthy periods of time. There is a need to balance the educational and social provision of schooling against the risk of spreading COVID-19. Simple models exist to quantify airborne infection risk but they are sensitive to parameterisation of a) the environment, b) the duration of exposure, and c) the disease – we identify routes to mitigate the uncertainties from two of these factors and assess the sensitivity of our results to COVID-19 parameterisation.

Usually, primary school children attend the same classroom day-in-day-out, and, assuming symptomatic individuals will not attend, the likelihood of infection within a classroom can be investigated by examining exposure over an entire asymptomatic infectious period (5-7 days for COVID-19). By sourcing existing monitored CO2 from classrooms around the UK, we directly assess the key environmental conditions and develop models that allow for variable occupancy during the infectious period.

Our results show wide variation in the predicted number of secondary infections between the monitored classrooms, including between classrooms within the same school and with the same automated ventilation provision. Solely accounting for changes in the indoor environmental conditions, systematic seasonal variations are also observed: wintertime is expected to give rise to a two-fold increase in infections, which is likely to be compounded by several other factors.