Ex-post energy savings assessment methodologies using smart-meter data: the case study of a switch from direct electric heating to an air-to-air heat pump
Résumé
The need to electrify end uses to achieve carbon neutrality by 2050 and strengthen energy independence means that we need to seek out and evaluate the most efficient technologies. To limit these future increases in electricity consumption, existing consumptions must also be reduced. Thus, retrofitting direct electric space heaters with a heat pump is a dual opportunity to speed up the decarbonization of the French residential sector and to moderate both individual and overall electricity demand increase. We then use this case study to evaluate a range of evaluation methods.
To assess the energy savings achieved by the switch from direct electric heating to air-to-air heat pump, in the absence of information to set up a control group (difference in difference), we monitored the daily electricity consumption of a dwelling, via the existing smart meter during 2 periods: 3 years with direct electric heating (DE) before retrofit, and 2 years after the installation of the air-to-air heat pump (HP).
The daily meteorological data were used to assess the thermal gradient (consumption versus temperature). The aim is to capture the heat-sensitive part of the daily consumption (mainly linked to space heating).
Several methods have been used to estimate energy savings by comparing the before and after situations: Linear, quantile and partial least square regressions, Multivariate adaptive regression spline model, and Machine learning based on generalized additive model.
The impact of the energy savings estimated according to the methodology is assessed and the uncertainties that may exist within each method and between methods is presented. The conditions of use and advantages of each method are discussed as they apply to other smart meter consumption, such as gas and other renovation.
Domaines
Sciences de l'environnementOrigine | Fichiers produits par l'(les) auteur(s) |
---|