Modeling of a Multi-Tube Solar Reactor for Hydrogen Production at High Temperatures

The thermo chemical dissociation of zinc oxide is the first step in a two-step water splitting cycle to produce hydrogen. In previous work, the dissociation of zinc oxide has been investigated in a single-tube solar reactor that was enclosed in a sealed quartz tube and irradiated by concentrated sunlight. That reactor concept allowed the successful demonstration of the zinc oxide dissociation in the temperature range 1500–2100 K at the High-Flux Solar Furnace (HFSF) at NREL in Golden, CO. The current modeling work focuses on the further development and optimization of the previously tested reactor design. The commercial software FLUENT is used to run simulations in 2 and in 3 dimensions. The new design consists of a multitube configuration in a closed cavity with a small aperture. The inner cavity wall has a layer of a highly reflective material. Parametric simulation studies have been conducted in order to investigate the influence of parameters such as solar flux concentration, number of tubes, tube size, zinc oxide mass flow and particle size (0.06 and 1 micrometer) on the overall efficiency. Solar average flux concentrations of 3000 and 6000 suns have been chosen. The secondary concentrator currently used at NREL (∼2000 suns average flux) will have to be redesigned in order to achieve about 3000 suns average flux. Between 3 and 10 tubes that are circularly positioned in the cavity have been simulated. The zinc oxide mass flow ranges from 2–20 g/min per tube. In 2D, an energy sink equation is incorporated that accounts for the heating-up of the inert gas and the zinc oxide particles and for the heat of reaction. The heat of reaction is calculated using a set of Arrhenius-type kinetic parameters that were determined experimentally. The simulation results show that the flux concentration, the zinc oxide mass flow and the number of tubes have the most significant effect on the thermal efficiency of the reactor. Conclusions are: 1) A reaction efficiency (energy solely used by the chemical reaction divided by the solar input) of about 29% with a total efficiency of about 36% is achievable. 2) Complete conversion can be obtained at temperatures in the order of 2000 K. 3) In a design for NREL’s HFSF (∼8 kW solar input power) a total zinc oxide mass flow of about 30 g/min is optimal. 4) More and smaller tubes lead to higher efficiencies. In the coming months, a new reactor will be constructed and tested at NREL’s HFSF.