Verbesserung der Abbildungsgüte aggregierter Modelle zur Simulation und Optimierung komplexer Fernwärmenetze

During the past decade the application of model based tools for simulation and optimization has become common in the field of energy supply. Due to additional degrees of freedom it is quite likely, that corresponding models will get more and more complex. The calculation of short term planning horizons as well as the general solvability of these models might be impossible in many cases. Therefore, in the future, a simplified but still sufficient accurate mapping of these systems will be very important. In order to describe complex district heating networks with several energy producers and recognizing the possibility of heat storage within the pipeline system, this explanation provides the essential structural components and planning tasks required for an optimal system operation. The inspection of up to date modelling, optimization and aggregation concepts, regarding also the corresponding need for development completes the definition of this elaboration's context. In a next step a method will be developed, which for the first time enables a high simplification of the district heating network model for not steady state operation. This strategy guarantees also a high correlation between the original and the aggregated network model. Differing from currently known aggregation methods for quasi-steady state systems, in this publication also the variables of each network element and each time step of the optimization horizon are taken into account when calculating the optimally aggregated district heating network model. When determining the optimal aggregated model of the district heating system, a multicritical objective function enables to adjust the model towards a defined scope of application. For this purpose the objective function contains appropriate weighting factors. The results of this publication indicate, that network models optimized by the described method provide a significant higher accuracy comparing to existing aggregation methods. The fields of application can be the calculation of complex scenario based simulations and optimization problems which are up to know unsolvable