Further development and validation of the "PROFet" energy demand load profiles estimator
Kamilla Heimar Andersen, Synne Krekling Lien, Karen Byskov Lindberg, Harald Taxt Walnum, Igor Sartori
Abstract
Long-term forecasts of the aggregate energy load profiles are crucial for energy system planning. Previous work has developed a load profile model named PROFet to forecast aggregated weather-dependent load profiles. This with an hourly resolution for buildings based on energy measurements from buildings connected to district heating. In this study, further development of the PROFet model is presented, including the extension of measurements in the database, the separation of building categories into three efficiency levels, and the separation of energy load profiles for electric loads, space heating, and domestic hot water heating. Load profiles created with PROFet for apartments have been validated against out-of-sample data (load measurements from buildings with hourly resolutions). This study shows that the out-ofsample datasets are in good agreement with the PROFet prediction. Also, the validation results are satisfactory as they are below the threshold values for the statistical indicators. A lasting contribution is a tool for creating aggregated load forecasts, which provides crucial information in grid planning and modelling. The results also provide typical load profiles that can be used as reference values when evaluating a neighbourhood/area's energy flexibility.