Date

8-1-2018

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Mechanical Engineering

First Adviser

Alparslan Oztekin

Abstract

Themochemmical energy storage (TCES) compared to other energy storage solution is more suitable due to possibility of storage at higher density and even higher temperatures. Thermal energy is stored/released using a reversible endothermic/exothermic reaction. For large-scale high temperature heat storage, TCES in nearly the only solution compared to the sensible or latent heat storage. However, high temperature TCES is at research stage and has not been commercialized yet. Among several potential reversible chemical reaction, metal oxide reduction/re-oxidation (redox) reaction have several advantages. They usually have high reaction temperature with high enthalpy of reaction, which make the thermochemical energy storage a better fit for large-scale applications. In metal redox reaction, air can be used as both heat transfer fluid and reactant. Using air will simplify the system considerably as the HTF storage system can be eliminated and air flow can be directly used in reactor eliminating any required intermediate heat exchangers. For new generation of large scale Concentrated Solar Power (CSP) plants, the thermochemical energy storage operating at high temperature is a tailor made for design requirements. Cobalt and manganese oxide are among the promising candidates for high temperature storage redox systems. In this study, the packed bed reactor has been selected as a storage reactor for redox system at high temperature. Cobalt oxide and Iron-doped manganese oxide are selected as redox reactant for the packed bed reactor. A mathematical model is developed including available reaction kinetics and mass, momentum and energy transport within the reaction bed and heat transfer fluid. Equations governing mass, momentum, and heat transport in the bed with the reaction kinetics are solved using finite element method. A simple packed bed reactor is selected as a base system for storage simulation study. Complete storage cycle simulation together with parametric study was implemented for two-dimensional model. Developed model is used for application into other different design configurations. To address main drawbacks of simple packed bed reactors, a modified split flow design configuration is proposed in which the HTF flow is split into direct and indirect contact mode. Parallel flow configuration of proposed modified split flow reactor is simulated. It was shown that flow splitting will reduce pressure drop along the reaction bed considerably while increasing the storage duration. It was shown that gained energy saving for pressure reduction is totally justified compared to the reduction in overall storage performance. Split design provides more control over the bed properties variations during cyclic storage operation and also the reactor is more stable in variation of bed properties after several successive storage cycles.

Experimental analysis was conducted on pure cobalt and manganese oxides to characterize the reaction kinetics. Pure manganese oxide passing two successive reduction during storage half cycle in which only the second one is partly reversible. Thus only modified systems such as doped type of manganese oxide can be applied for TCES application. For cobalt oxide system, it was shown that even pure sample can complete storage cycle and the reaction was complete for considered heating and cooling rate. There was a shifting and transition detected at initial cycles for the discharge half-cycle. The system behaved more stable condition after passing initial cycles. Morphological analysis was conducted on the pure sample after a few initial cycles to better understanding the detected transition. It was shown using SEM results that there was considerable grain growth observed after few cycles. In order to differentiate the effect of reaction and temperature for transient condition and considerable grain growth, temperature analysis conducted via putting sample at certain temperatures below and above reaction temperature. It was shown that the grain size growth occurred mostly at 1000 C and the effect of reaction on the grain size was not strong. It was inferred that the detected transient behavior in reaction at later cycles could be due to grain size variation at high temperature.

Share

COinS