Impact of natural ventilation on mitigating airborne disease transmission risks in office rooms: Tool development

Abstract

Ventilation is crucial for indoor air quality and controlling the spread of airborne diseases. However, designers face challenges in evaluating ventilation strategies early in the design process because current tools are complicated and require extensive input data and specialized knowledge. Our study aimed to develop a methodology for evaluating natural ventilation strategies in the initial stages of building design, with a focus on reducing the transmission of airborne disease, managing CO₂ concentrations and maintaining thermal comfort. We first identified the maximum feasible hours for natural ventilation throughout the year using weather data and constraints. Then, we modelled 256 ventilation scenarios for an office room in Tehran, simulating each under five modes that varied by exposure time, mask use and activity level with the Climate Studio Plugin. We generated 1280 ventilation modes and built predictive models using ensemble learning. The Random Forest model (R0, fifth scenario) and the XGBoost model (comfort hours) achieved the highest accuracy (R² = 0.996) for annual index prediction. The cross-ventilation-NS configuration (north-south) yielded the best outcomes in this study, underscoring the importance of strategic ventilation design to promote health and safety. Finally, we transformed this data into a user-friendly tool, developed using the Streamlit program.