Comparison between Direct and Indirect ORCs for CSP Applications

Concentrating solar power is still a niche market for ORCs counting several small applications but only few large scale plants. The main limit of solar ORCs with respect to PV systems is related to the complexity of this technology and the consequent higher specific cost. However, the possibility to adopt a thermal energy storage (TES) allows solar ORCs to providing dispatchable solar energy and to be possibly competitive with large PV systems integrated with electrochemical storage. The aim of this work is to define the best solar ORC power plant configuration for different maximum solar field temperatures by comparing conventional indirect power plants against non-conventional direct systems. Conventional indirect cycles consists of a solar field where the heat transfer fluid is heated up to maximum around 400°C, hot oil can be stored in the TES or used directly as heat input for the ORC generally designed as a subcritical saturated cycle. On the contrary, in direct cycles the working fluid flows directly into the solar field and stored in liquid phase. Hot working fluid is then throttled (partially or completely) and vapor fraction is expanded in a turbine. This configuration allows avoiding the HTF/working fluid heat exchanger possibly leading to a lower investment cost and higher efficiency. A Matlab code have been implemented in order to carry out the thermodynamic optimization of both indirect and direct solar ORCs. Indirect systems are optimized varying the evaporation and condensation pressure and the turbine inlet temperature while direct systems by varying both the condensation and the flash pressure. Both recuperative and non-recuperative configurations are investigated considering a large number of working fluids from Refprop database. The optimal working fluid is selected considering the system efficiency as figure of merit but optimal power plant selection considers also the size of the storage system. Respect to the scientific literature some novel aspects are implemented: (i) a model of the solar field in order to link pressure drops to fluid properties, (ii) a turbine efficiency dependent on the expansion volume ratio and (iii) an extensive sensibility analysis on different assumptions like minimum pressure and components performance.