Regression analysis of the energy consumption of tertiary buildings

Energy signature methods are applied over three tertiary buildings in the UK, Sweden and Spain, based on both simulations and experimental data, for pre- and post-retrofit scenarios. Variations in their energy profiles relate to differences in climate severity, usage pattern (continuous/discontinuous) and HVAC scheduling. This study discusses the impact of such particularities for obtaining a steady-state linear regression model of the dependence of heating energy load against climate data. The choices of dataset and time step have important implications for the results obtained.