Reliability and Performance Gap of Whole-Building Energy Software Tools in Modelling Double Skin Façades
MetadataShow full item record
The careful design of the façade is one of the most influential strategies to lower the energy use in a building. A double skin façade (DSF) is one type of façade that allows the interaction between the outdoor and the indoor environment to be managed in a more advanced way, by increasing the control over the energy transfer between the two environments, while providing high architectural flexibility and transparency. The design of the thermophysical performance of a DSF is a complicated process that has to take into account several aspects, such as geometric parameters, thermal properties, ventilation strategy, shading devices, and the integration between the façade and the building energy concept. There exist different whole building energy software tools (BEST) that practitioners can use to predict the energy and indoor environmental performance of a building and to support an informed choice to select the most appropriate building components during the design phase. However, when it comes to the simulation of DSF in BEST, complexity and inaccuracies in prediction usually rise, as these envelope systems are characterised by a thermophysical behaviour that requires a more advanced modelling than the possibilities conventionally embedded in BEST. This paper reviews the scientific literature to show evidence on how BEST are used to predict the thermophysical behaviour of DSF, together with reporting the existing modelling capabilities for some selected BEST. The purpose is to highlight the challenges associated with the modelling of DSFs and to identify the major gaps between measured performance and prediction though BEST. The findings indicate that gaps are mostly connected to the dynamic behaviour of the DSFs and in particular the airflow within the façade cavity. The challenges associated with the modelling and simulation for each software tool, and the skills necessary to recognise and implement the best-suited model among the different options available are also discussed.