Elsevier

Energy

Volume 29, Issues 5–6, April–May 2004, Pages 771-787
Energy

A two-cavity reactor for solar chemical processes: heat transfer model and application to carbothermic reduction of ZnO

https://doi.org/10.1016/S0360-5442(03)00183-XGet rights and content

Abstract

A 5 kW two-cavity beam down reactor for the solar thermal decomposition of ZnO with solid carbon has been developed and tested in a solar furnace. Initial exploratory experiments show that it operates with a solar to chemical energy conversion efficiency of about 15% when the solar flux entering the reactor is 1300 kW/m2, resulting in a reaction chamber temperature of about 1500 K. The solid products have a purity of nearly 100% Zn. Furthermore, the reactor has been described by a numerical model that combines radiant and conduction heat transfer with the decomposition kinetics of the ZnO–carbon reaction. The model is based on the radiosity exchange method. For a given solar input, the model estimates cavity temperatures, Zn production ra4tes, and the solar to chemical energy conversion efficiency. The model currently makes use of two parameters which are determined from the experimental results: conduction heat transfer through the reactor walls enters the model as a lumped term that reflects the conduction loss during the experiments, and the rate of the chemical reaction includes an experimentally determined term that reflects the effective amount of ZnO and CO participating in the reactor. The model output matches well the experimentally determined cavity temperatures. It suggests that reactors built with this two-cavity concept already on this small scale can reach efficiencies exceeding 25%, if operated with a higher solar flux or if one can reduce conduction heat losses through better insulation and if one can maintain or improve the effective amount of ZnO and CO that participates in the reaction.

Introduction

A number of papers describe the use of concentrated solar energy as a source of process heat for the high temperature carbothermal reduction of ZnO [1], [2], [3], [4]. By substituting solar energy for fossil fuels or for electricity, these articles show that a Zn producer could reduce its CO2 emissions by a factor of 5–10 [1]. If the carbon source is biomass, one combines a human produced solar chemistry process with a solar process given to us by nature--photosynthesis, to create a potentially CO2 neutral Zn plant. Some of the articles also indicate that Zn can be used to split water to produce H2 [5] and that one can view Zn as a fuel to a fuel cell or battery [6]. From this point of view, the solar process is a means of storing sunlight in the form of a chemical fuel: the Zn exiting a solar reactor is fed to a fuel cell for electricity production, and then the ZnO produced in the cell is recycled to the solar furnace. A solar thermal carbothermal process is thus a potential sustainable path for producing an important commodity or electricity.

We recognize that the sustainable potential of the process depends, in part, on economics. A solar thermal process for producing Zn will be industrially interesting if its cost per unit of CO2 reduction is lower than that of any other process that leads to similar or better CO2 mitigation levels. Because the heliostat field can represent up to 40–50% of the initial capital cost of a solar plant [1], it is important that the solar chemical reactor effectively uses the sunlight to effect the desired chemical reaction; the more efficiently the reactor uses sunlight to produce Zn at a given rate, the smaller the required heliostat field and consequently the lower the capital cost of the plant. Thus, the industrial potential of the solar process is strongly tied to our ability to design, build, and run an efficient reactor.

We have argued that an industrial solar reactor must respect the transitory nature of solar en ergy [7]. This constraint implies that the reaction proceeds with minimal delay when the sun is available. It withstands the thermal shocks that occur under the severe environment of transient intense solar radiation. The reactor is designed such that a large fraction of the solar energy entering the reactor drives the thermal chemical reaction: the radiant energy streaming through the reactor matches the rate of the desired solar reaction so that one minimizes the fraction of the solar energy that ultimately leaves the reactor via radiation or convection heat transfer.

In this paper, we describe part of our efforts at achieving this last objective. Specifically, we present methods and latest findings as they pertain to the design of a solar thermal chemical re actor that we use to effect the reactionZnO+CsolidZn+CO.

Because the design process is multifaceted and iterative in nature, it is necessary for us to fo cus on the details of only a portion of the design process. Our focus, here, is on a heat transfer model that combines the reaction kinetics of the carbothermal reaction with radiation and conduction heat transfer. Furthermore, we illustrate how we combine the model results with experimental results to judge the reactor’s design, and then how this information is used to further improve the reactor.

Section snippets

The two-cavity reactor

The two-cavity reactor shown in Fig. 1 evolved from a design described in Ref. [8]. This original work developed the mechanical and global heat transfer arguments behind the design concept. Here we simply describe important features of the reactor that help illustrate how it works, so that one can understand the significance of the results coming from our latest numerical model and experimental program.

Concentrated sunlight enters the reactor at its top. It then passes through a water cooled

The numerical model of the reactor

The reactor is modeled under stationary heat transfer conditions. Wall losses and the energy consumption for the endothermic ZnO-reduction enter the model as boundary conditions on the respective wall surfaces. The heat transfer model is based on a method for solving the radiant interchange among diffusely emitting and diffusely reflecting gray surfaces within an enclosure. The model further presumes that each surface of the enclosure is isothermal and that the radiosity, which is the rate at

Experiments

Two types of exploratory experiments were performed with the two-cavity reactor shown in Fig. 1:

  • 1.

    Batch experiments, in which 300 g of ZnO/C-mixture were put in the reactor prior to heating. An approximately steady state with regard to the inner temperatures and gas flows was reached after some 10–20 min. The total duration of an experiment was 40–60 min.

  • 2.

    Experiments with continuous feeding: only after the reactor was heated to a temperature of around 1200 C in the lower (reaction) cavity, a screw

Numerical analysis and comparative discussion with experimental results

Our numerical model described in Section 3 simulates the solar experiments. In order to solve the heat transfer model, we need to specify the reactor’s wall losses as a function of the wall’s temperature. For this purpose, one could formulate a description of these losses on surfaces 2, 3, 4, 5, and 7 based on the actual wall materials, their thickness and properties, as well as the convection situation outside the walls. Our objective here, however, is to show a general nu merical reactor

Conclusions

A 5 kW two-cavity beam down reactor for the solar thermal decomposition of ZnO with solid carbon has been developed and tested in a solar furnace. Although the reactor has not yet gone through an optimization process, initial exploratory experiments show that it operates with a solar to chemical energy conversion efficiency of about 15% when the solar flux at the aperture is near 1300 kW/m2. The solid products have a purity of nearly 100% Zn. The reactor operates in a batch or continuous feed

Acknowledgements

Financial support from the BFE—Swiss Federal Office of Energy—is gratefully acknowledged. We thank A. Berman for fruitful discussions concerning the ZnO–C-kinetics and S. Kräupl, M. Brack, P. Häberling, B. Schaffner, A. Steinfeld and D. Wuillemin for discussions and their help prior to and during the experiments. Experiments were performed at the Solar Furnace, Paul Scherrer Institute, Villigen, Switzerland.

References (19)

There are more references available in the full text version of this article.

Cited by (57)

View all citing articles on Scopus
View full text