Bottom-up statistical analysis of the energy consumption of French single-family dwellings
Résumé
Implementing effective energy policies in the residential sector requires better understanding of the sources for the dispersion of energy consumption amongst households. Bottom-up statistical models have been identified as one major modelling technique, particularly accounting for the diversity of inhabitants behaviours. In the various statistical models of the literature, behaviour characteristics are seldom incorporated in the data set but socioeconomic data are most often used as " proxy " of occupant behaviour. The present study is based on a detailed survey, combining energy billing data with technical, geographical, socioeconomic and behavioural variables. The corresponding sample, despite of its limited size (420 individuals), is representative of French households. A statistical model relating energy consumption to the other variables has been applied, enabling simultaneous use of quantitative and qualitative explanatory factors (ANCOVA, analysis of covariance). The main determinants found for energy consumption in the sector of single-family dwellings in France are, by decreasing order of weights, surface area, type of main heating system, age of the household head and climate zone. We detected that the most influential behaviour variable is night temperature setting reduction. Explanation and prediction capacities of the model as the accuracy of the model coefficients are studied and some possibilities of improvement are proposed.
Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...