Uncertainty handling control algorithms for demand response with modulating electric heating devices

The flexibility of electric heating devices coupled with thermal storage can help to cope with the increasing share of volatile renewable energy sources in the electricity grid. Scheduling-based demand response approaches for optimally exploiting these flexibilities use demand and generation predictions to calculate an operative schedule of the heating devices. Due to deviations between predicted and real energy profiles, additional uncertainty handling methods are essential which adjust the actions imposed by the schedule to the current situation. In this paper, we introduce corrective control algorithms for buildings in smart grids with modulating heating devices that can compensate the uncertainties of predictions. The results show that our developed approaches avoid violations of the inhabitants' comfort limits and decrease the surplus energy (and thus increase the self-consumption rate) of photovoltaic systems compared to a conventional control strategy. Further, our analysis reveals that uncertainties affect the load shifting potentials of electric heating devices and lead to increased surplus energy.