A review and an analysis of the residential electric load curve models

International audienceDue to the growth of electric end-uses, the management of the variations in time of the electric power demand has become essential, especially in the residential sector. According to this issue, the anticipation of the power demand is of great interest. This implies a better knowledge of the electric load curve of the household stock. Papers about understanding and forecasting energy demand are numerous but studies on building's load curves are rare. In this paper we propose a cross analysis of some existing methods capable of building up a residential electric load curve. Two main types of load curve models have been identified in the literature: top-down and bottom-up methods. Even if the review presents two existing top-down approaches, the authors focused the further analysis on bottom-up models. For each of them we first identify its functional characteristics: finality and scope, input data required, output format, modeled appliances and end-uses covered, generation of the diversity and validation of the model. Secondly, we establish a bloc diagram representing its architecture with focus on the mathematical model chosen. Finally, the authors list the limits of the model in view of the criteria needed to build up an ideal, bottom-up and technically explicit load curve model for the residential sector. A cross reading of the different methodologies is proposed through a global table that characterizes and sums up the analyzed models. Moreover, a graphical representation of the models studied is proposed, according to three criteria (range of application, modelling of the diversity, time scale accuracy) that allows us to compare them "at a glance". To conclude, an identification of the gaps among existing models is proposed. It consists of listing the end-uses, the appliances and other behavioural correlations which may affect the load curve and that are not included in the reviewed methodologies and need further research