Generalization Approach for Models of Thermal Buffer Storages in Predictive Control Strategies

The demand for flexibility in a future electricity system, mainly based on renewable energies, will increase. Flexible building devices combined with predictive control strategies can help to cover this increasing demand for flexibility. Predictive control strategies allow an optimization of the operation mode of e.g. heat pumps or combined heat and power units with a regard to the current availability of electricity in the power grid. With thermal storage units, the available flexibility of the thermal generation units increases and avoids influences on the thermal comfort in the building. For an optimized operation, a supervisory control has to be equipped with models of the building devices and the storage units. Those models have to be as simple as possible to reduce computation time as well as the effort during the commissioning process. This paper describes a generalization approach for a model of thermal buffer storages. The generalized model is derived from a detailed model, developed in Dymola, and is based on characteristic curves, which describe the temperature of the storage medium depending on the current state of charge. The developed models are validated using measurement data from a hardware test bench. The results of the validation show, that the deviation of the simulated storage temperature are between 5 and 22 %, depending on the observed temperature layer in the current operation mode (loading or unloading).