Grey-box modelling for naturally ventilated buildings

Ghjuvan Antone Faggianelli (1) , Adrien Brun (2, 3) , Etienne Wurtz (2, 3) , Marc Muselli (1)

1 SPE - Sciences pour l'environnement
2 CEA, LITEN, Service bâtiment et systèmes thermiques, Le Bourget du Lac, France
3 INES - Institut National de L'Energie Solaire

Among passive strategies to reduce energy consumption in buildings, we focus on natural ventilation, which can bring an important decrease in temperature during summer depending on climate. Despite its simplicity , it needs particular attention to be efficient and can be improved with building control. In this paper, we focus on a simplified thermal model based on an electrical analogy (6R2C), coupled with a statistical airflow model and calibrated for a residential building in Mediterranean climate. A full-scale experiment on this building, located in a coastal area of Corsica, allows to build and test models adapted to the case study. Both simulations and measurements are used to calibrate and test the simplified model. It is shown that the calibration phase has a great impact on model performance. An adapted calibration thus allows to reach very low errors, far bellow 0.5° C on average, during the whole summer period. An application of this model is also proposed to control night ventilation in order to limit the number of operating cycles (windows opening and closing) and avoid over-cooling.

Domains: Engineering Sciences [physics]

Ghjuvan Antone Faggianelli, Adrien Brun, Etienne Wurtz, Marc Muselli. Grey-box modelling for naturally ventilated buildings. 14th Conference of International Building Performance Simulation Association, Dec 2015, Hyderabad, India. pp.720-727. ⟨cea-01615331⟩