Dynamic modeling of a steam Rankine Cycle for concentrated solar power applications
Tutor / director / avaluadorQuoilin, Sylvain
Tipus de documentTreball Final de Grau
Condicions d'accésAccés obert
Concentrated solar power (CSP) is expected to play a key role in the necessary energy transition towards more sustainability. However, this type of system is inherently subject to transient boundary conditions such as varying solar irradiation. Therefore, advanced control strategies are required to maintain safe operating conditions and to maximize power generation. In order to define, implement and test these control strategies, dynamic models of the system must be developed. This thesis aims at developing a model of a steam Rankine Cycle coupled to a field of parabolic troughs. The modeled system does not correspond to an existing plant, but its characteristics are defined as realistically as possible with information coming from different sources. Simplified but also physical, lumped dynamic models of each component (boiler, turbines, condenser, solar collectors) have been developed and parametrized using the ThermoCycle library, written in the Modelica language. These models have been further interconnected to build the CSP plant model, whose response has been tested to fluctuating atmospheric conditions. The proposed library of models is based on an innovative lumped-parameter approach aiming at developing physical models that are significantly more robust and computationally efficient than the traditional libraries of models already available. The final purpose of these models is high level simulations (e.g. for control purposes), but not the modeling of detailed physical phenomena. The different models have been successfully tested with the example of the CSP plant, but can also be applied in other fields of thermal engineering. They proved to be more robust and much faster than the traditional models, which was the objective. However, in the scope of this work, it has not been possible to validate them with experimental data or with more detailed models. This should be the priority for future works.