Abstract
This paper presents a case study of a recently built high-performance Canadian social housing building with the aim of comparing the expected and measured energy consumptions and to identify the parameters affecting the most the energy need. A monitoring system compiles at a 10-min frequency information related to the energy use and the thermal conditions observed in the building and its HVAC system. The building has the particularity of comprising two symmetric sections made of different timber structure systems. No significant differences of energy consumption were detected between the two parts of the buildings. However, a large variance was observed when comparing each dwelling individually regardless of their structures. The orientation of the dwelling also exhibited a minimal influence compared to these variations, suggesting that occupant behavior is the dominant factor explaining dwelling-to-dwelling variability and is thus critical for understanding energy use in residential buildings. Regression analysis showed that specific occupant actions, such as opening windows in winter or using electrical appliances, have a great impact on the energy balance of the apartments. In 2016, the performance gap between measured and expected total energy demand of the building was 74%. With the use of the large dataset coming from the building, it was possible to determine the causes behind this large gap for the reference building.