Abstract
For realistically modelling the residential heat load request for space heating and hot water the authors define a a suitable two step stochastic model. The proposed methodology allows to predict the daily heat load profiles and the corresponding confidence intervals during the heating season.