Understanding the reasons behind the energy performance gap of an energy-efficient building, through a probabilistic approach and on-site measurements

The performance gap, defined as the difference between the measured and the calculated performance of energy-efficient buildings, has long been identified as a major issue in the building domain. The present study aims to better understand the performance gap in high-energy performance buildings in Switzerland, in an ex-post evaluation. For an energy-efficient building, the measured heating demand, collected through a four-year measurement campaign was compared to the calculated one and the results showed that the latter underestimates the real heating demand by a factor of two. As a way to reduce the performance gap, a probabilistic framework was proposed so that the different uncertainties of the model could be considered. By comparing the mean of the probabilistic heating demand to the measured one, it was shown that the performance gap was between 20–30% for the examined period. Through a sensitivity analysis, the active air flow and the shading factor were identified as the most influential parameters on the uncertainty of the heating demand, meaning that their wrong adjustment, in reality, or in the simulations, would increase the performance gap