Enabling large-scale dynamic simulations and reducing model complexity of district heating and cooling systems by aggregation

District heating and cooling (DHC) systems are considered cornerstones of a future heat and cold supply due to their ability to integrate renewable and waste heat sources as well as long and short-term storage technologies and their flexibility for integration of other infrastructures. Therefore, DHC systems may represent the central hub of an interlinked overall energy system. These options nevertheless lead to a more complex system if implemented as the number of technical components and potential interactions increase and as the energy demands on the network grow. The transportation of the heat supplied to the consumers through water flow involves both time and temperature-dependent changes based on the mass flow rates and the heat losses to the surrounding from the pipes. The dynamic properties of the consumers and the distribution pipes strongly influence the operation of the district heating systems. Dynamic modeling tools are required to cope with the increasing complexities and to optimize the operation of the DHC systems by using the knowledge of time delay in the heat transport, temperature distribution and flow characteristics. Nevertheless, it is computationally challenging to simulate large-scale DHC networks dynamically. One proposed method for improvement in this context is aggregation, which simplifies the topological complexities of the original network by reducing the number of pipe junctions (branches) and pipes. This paper focuses on the analysis and evaluation of the application of two aggregation methods, the Danish and the German method, in a dynamic modelling environment and comparing the aggregation to the monitoring data of a virtual district heating network with 146 consumers and an existing district heating network with 66 consumers. These evaluations are based on comparison of the simulation results of the aggregated network with fewer consumers than the original network in terms of accuracy, information loss and computational time. The results show that the CPU time in the simulation of this network can speed up approximately more than 95% by aggregating the network down to 3 consumers without significant information loss. We will highlight the current potentials and limitations of both aggregation methods in terms of integrating them in the planning, modelling and operation of DHC systems, including the 4th generation district heating systems.