A two-level optimisation approach for underground natural ventilation based on CFD and building energy simulations

Abstract

Natural ventilation is a historical and promising passive design for conserving building energy. However, a noticeable scarcity exists in design strategies and academic studies on underground natural ventilation. This study introduces a two-level optimisation approach for underground natural ventilation, considering form and environmental characteristics. At the initial level, computational fluid dynamics simulations were employed to quantify and visualise the momentary ventilation performances of various passive strategies. These simulations also furnished precise wind pressure coefficients. Subsequently, at the second level, this study predicted the annual energy savings of the proposed prototype by integrating an airflow network model with building energy simulations. To adapt design parameters, sensitivity analysis and multi-objective optimisation techniques were applied. These numerical models underwent validation against measured data from analogous studies. The results reveal that the ‘inlet buried pipe + outlet shaft’ led to a 6.9 % reduction in energy consumption for heating, ventilation, and air-conditioning in underground hybrid ventilation compared to underground mechanical ventilation. Moreover, it achieved a 28.8 % decrease compared to aboveground mechanical ventilation. Optimising air temperatures following climate, soil conditions, and building characteristics is imperative, aligning with design standards. This study contributes a quantitative and visual approach to support underground natural ventilation and addresses the limitations of empirical designs, providing a foundation for future research and practical applications.