Modellierung von solaren Partikelreceivern mit der Diskreten Elemente Methode

Abstract

Particles are envisaged as a redox material in solar-thermochemical fuel production processes and as well as a heat transfer medium in solar thermal power plants. They are heated in solar particle receivers, which have mostly been evaluated with continuum models so far, even though the Discrete Element Method (DEM) usually describes the particle motion more accurately. Reasons are the lack of heat transfer models needed for a particle receiver, missing contact model parameters for potential particle types proposed for solar receivers and little experience with the method in the solar thermal research community. In the present work these hurdles are addressed, with the ultimate aim that the method will be part of the methodology toolbox for future researchers. Several models were developed: one for the conductive heat transfer through the contact point and void space between two particles, one for the same heat transfer modes between a particle and a wall, a radiation model based on Monte Carlo ray tracing, a wall conduction model and a model for the chemical reduction of ceria, which occurs in concepts for solar-thermochemical fuel production. The inter-particle conduction model includes a pressure dependence as it is based on the extended Zehner-Bauer-Schlünder model for the thermal conductivity of packed beds. The applicability of this continuum model under vacuum and high temperature conditions was validated in a vacuum experiment in a solar simulator. Also the other models were successfully compared against solutions of various test cases. In context of the model developments, a critical time step limit for the common particle temperature updating process in the DEM was derived for the first time and proven to be reasonable in a stability test. Additionally, contact parameter sets for five particle types envisaged in solar receivers have been determined by a custom calibration approach. It is based on five bulk experiments, which are used to calibrate the contact parameters in three stages. More precisely, the DEM models of the respective experiments in each stage are described by surrogate (Kriging) functions, whose inputs are then optimized to match the experiments and thus find the contact parameters. The calibration was also performed for increased particle diameters to provide parameters for coarse-grained and therefore faster simulations. Finally, the application of the models as a whole is demonstrated by a simulation of the prototype CentRec particle receiver. Mass flow fluctuations observed in experiments could be reproduced and were analyzed in detail. The simulated particle outlet temperature and the receiver efficiency were in good agreement with their experimental counterparts. In summary, the DEM has been shown to be a very useful method for the analysis and the design of solar particle receivers and should be used for this purpose in the future.