Scale-bridging within a complex model hierarchy for investigation of a metal-fueled circular energy economy by use of Bayesian model calibration with model error quantification

Metal energy carriers have recently gained growing interest in research as a promising storage and transport material for renewable electricity. Within the development of a metal-fueled circular energy economy, research involves a model hierarchy spanning from micro to macro scales, making the transfer of information among different levels of complexity a crucial task for the implementation of the new technology. Chemical reactor networks (CRNs) are reactor models of reduced complexity and a promising approach to accomplish the scale-bridging task. These models combine the ability to incorporate detailed chemistry with the efficiency to perform large parameter studies. This holds if valid information from CRNs can be obtained on a much denser set of operating conditions than is available from experiments and elaborated simulation methods like Computational Fluid Dynamics (CFD). An approach for CRN calibration from recent literature, including model error quantification, is further developed to construct a CRN model of a laboratory reactor for flash ironmaking, using data from the literature. By introducing a meta model of a CRN parameter, a simple CRN model on an extended set of operating conditions has successfully been calibrated and shown to perform well. A performance map has been computed from the calibrated model, revealing its computational tractability and enabling robust optimization. The study thus shows how CRNs can enhance the accuracy of reactor data in thermo-economic and optimization models on the power plant or full system scale. Additionally, the study has revealed several key steps for future research for the implementation of the scale-bridging workflow up to an industrial scale, which are comprehensively discussed.