Empirical and numerical approaches to predicting the thermal performance of green roofs

Abstract

Green roofs are increasingly recognised as a sustainable option for reducing building energy demand. However, given the large investment required for their construction, accurate modelling methods are needed to predict their economic benefits for building owners and maximize their effectiveness. The representation of vegetation in the green roof energy balance model literature is reviewed to identify their limitations. It is concluded that an overemphasis on single source models, minimal validation periods and limited input data is likely burdening the robustness and applicability of these models. Using data collected in Calgary, Halifax and London, this study then aims to develop empirical models to offer another approach for predicting the thermal performance of green roofs. Significant multiple linear regression models highlight the importance of net radiation, air-to-surface temperature difference and humidity in predicting the substrate heat flux, with the green roof in the driest climate; Calgary, being the most effective.