Background: Pupils and teachers spend up to a third of their day in primary schools, making air quality in classrooms a crucial determinant of their health. Previous studies have shown that carbon dioxide (CO2) levels in Slovenian classrooms often exceed the recommended and permissible levels. These classrooms primarily rely on natural ventilation, the effectiveness of which depends on various factors. Therefore, we aimed to develop a tool to predict classroom air quality and assess the CO2 exposure of pupils and teachers.
Aims and hypotheses: We aimed to develop a CO2 prediction model and analyse its suitability for monitoring air quality trends in Slovenian classrooms across different seasons. Additionally, we wanted to analyse the effectiveness of natural ventilation on children's exposure to air pollutants in classrooms. We set two hypotheses: (1) A validated mathematical multi-parametric model for predicting classroom CO2 levels will accurately predict classroom CO2 levels according to the classroom classification of indoor air quality. A simultaneous evaluation will confirm the feasibility of incorporating parameters specific to the indoor environment and the classroom activities into the CO2 prediction model. (2) The variation in the values of selected chemical, physical and microbiological pollutants in the classroom air, primarily sourced from the indoor environment, will show a statistically significantly similarity to the variation in CO2 levels in classroom air.
Methods: In the first phase, we conducted a cross-sectional study on air quality and natural ventilation in primary schools in Slovenia. Based on the identified factors affecting classroom air quality, we developed and content-validated a questionnaire. In the second phase, we developed a model to predict CO2 concentrations, considering the technical characteristics of the classrooms, the indoor and outdoor environment, the number of people in the classroom, CO2 generation, and ventilation. Model was validated using CO2 measurements taken in the classrooms during school hours. We also evaluated the model’s applicability for predicting variations in particulate matter (PM10, PM2.5 and PM1) and VOC. In the third phase, the developed model was used to assess natural ventilation and CO2 exposure of pupils and teachers in Slovenian primary schools. We developed scenarios for natural ventilation in small, medium, and large classrooms during both the heating and non-heating seasons.
Results: The results of the first phase, a cross-sectional study, indicated that certain sources of pollutants and factors affecting classroom air quality in Slovenian primary schools are statistically significantly related to the school’s micro location and the year of construction. In the second phase, we developed a model to predict classroom CO2 levels, based on the technical characteristics of classrooms and processes in 3rd-grade classrooms in Slovenia. The CO2 levels predicted by the model showed a very strong and statistically significant correlation with the measured CO2 values. The model CO2 values have the most significant correlation with PM2.5 among the remaining monitored pollutants. The third phase results showed that the initial CO2 levels affect CO2 levels until the 20-minute break, after which this effect is nearly nullified by prolonged classroom ventilation. CO2 exposure above 1000 ppm, which impacts the health and well-being of pupils and teachers, is lower during the non-heating season and in small classroom. During the heating season, CO2 exposure above 1000 ppm is reduced in classrooms ventilated before school, whereas pre-school ventilation has no or very little effect on CO2 exposure above 1000 ppm during the non-heating season.
Conclusion: The first hypothesis is accepted, and the second hypothesis is partially accepted. During the heating season, pupils' and and teachers' exposure to CO2, reaches levels that impact health and well-being more than half of the time, regardless of the ventilation scenario. The exception is in small classrooms with more intensive natural ventilation (scenarios C and D).
|