Abstract

The development of thermal systems for supermarkets is a challenging task. Both, heating and cooling demands at different temperature levels have to be satisfied under individual boundary conditions. In combination with a broad range of available technologies and components, a high number of possible system layouts exist. Thus, various types of refrigeration systems can be found in supermarkets: Central refrigeration systems with one or two stages and direct evaporation, central systems with a secondary loop or systems with (semi)-plug-in-cabinets. The system topology and operating strategy depend on climate conditions, building scale, customer’s occupancy or evaluation criteria. In practice, established solutions based on experience are used. However, comparing all alternative concepts is difficult. Beside the consideration of investment costs, it is essential to evaluate the energy consumption. For the calculation of energy consumption, considering dynamic interactions between components is crucial.

To compare different system layouts under consideration of dynamic interactions, an optimal operating control has to be applied. Furthermore, the high number of possible topologies makes it necessary to reduce the complexity for the selection of components and their interconnections. Therefore, software based methods are needed to efficiently reduce complexity and evaluate system alternatives in a dynamic environment. This paper presents a procedure that supports the user to find an optimal system topology under individual conditions. As an example, a secondary-loop refrigeration system with low and medium temperature cabinets is applied.

The user defines ambient conditions and requirements such as cooling load and temperature setpoints. Additionally, a set of transient, non-linear models for available technical equipment is defined. The parametrized, ready-to-use models are managed in a catalogue platform. In the catalogue, additional information is stored, like valid operational ranges, which is used during optimization. On this information basis, an algorithm deduces a reasonable refrigeration system layout. Intermediate result is a ready-to-simulate system. It contains only catalogue models that have physical reasonable interconnections. Subsequently, the system’s fluid flow rate of each connection is optimized. The result of the optimization is used for evaluation of the system layout and further reduction of its topology. The paper shows, that using simple input information, the complexity of the optimization problem can be extremely reduced. The suggested procedure is capable to deploy an optimal system topology under consideration of non-linear dependencies.