Statistical study on the link between real energy use, official energy performance and inhabitants of low energy houses

Energy performance regulations are becoming increasingly strict and governments supply simplified calculation tools to assess whether new buildings fulfil the requirements. However, one can wonder what the accuracy of those tools is for assessing the next generation of houses, that will have to fulfil the upcoming energy requirements. In order to investigate the discrepancy between predicted and real energy use in low energy houses, 537 dwellings were analysed. Data on building characteristics and theoretical energy use from the Flemish EPBD-database was complemented with data from the energy utilities and a survey of the inhabiting households, providing information about the households, their user behaviour and real energy use. While an undeniable correlation was found between theoretical and real energy use, the EPBD-method overestimated the heating energy use for most of the cases. Two building related parameters and two user related parameters proved to have a significant impact on that gap: the use of default values for the air tightness of the envelop and for the efficiency of the gas boiler, the heating profiles of the master bedrooms and the amount of baths and showers taken by the inhabitants. However, two comments must be made. First, the dataset consists of early adopters who could afford such energy performance years before it would be imposed and are therefore not representative of the average household. In addition, the analysis showed significant correlations between household characteristics on the one hand and building characteristics and performance on the other. These last two points question the possibility to extrapolate findings from samples of existing forerunners towards prognoses on future, entire building stock level.