Master's Thesis Dominik Hering


Optimization of the dimensioning of grid-connected heating systems

Hering Copyright: EBC Generation of control signal for a CHP unit, using a residual load profil. Release signal 2 indicates forced operation, release signal 1 normal operation and release signal 0 minimal operation.

In the course of the german "Energiewende" the electrical grid is faced with a shift of power generation from centralized power plants towards a more decentralized supply of electrical power from renewable energy sources. The volatile nature of renewable sources faces the low- and medium voltage networks with an increasing flexibility demand. Concurrently the distribution of grid-connected building energy systems, such as combined heat and power or heat pumps is promoted. Both technologies offer the potential to offer eletrical grid services by advantageous scheduling.

This work investigates the dimensioning of building energy systems. Depending on the relative dimensioning of thermal energy storages and heating devices, the building energy system can be operated more or less flexible. To optimize the building energy system dimensioning, an evolutionary optimization algorithm is implemented. The evolutionary algorithm uses simulation results to evaluate and compare different dimensionings. This results in a computational expensive evaluation function, causing the optimization process to take a long time. Therefore a pilot test is executed, based on design of experiment methods, to find a good set of parameters for the algorithm to minimize calculation time.

Finally, the building energy systems are optimized regarding their flexibility potential and their additional cost. The optimization results are then examined towards their sensitivity of the selected control strategy and the grid characteristics.