Rigorous Model and Experimental Verification of the Radiation Field in a Flat Plate Solar Collector Simulator Employed for Photocatalytic Reactions

BRANDI, R. J.; ALFANO, O. M.; CASSANO, A. E.

CHEMICAL ENGINEERING SCIENCE

PERGAMON-ELSEVIER SCIENCE LTD

Año: 1999 p. 2817 - 2827

0009-2509

In any kinetically controlled photocatalytic process the catalyst activation is always a photochemical act that depends upon the local value of the volumetric rate of energy absorption (LVREA). In a heterogeneous photocatalytic reaction, a precise evaluation of the LVREA can be obtained when the spatial and directional distributions of radiation intensities are known. With this purpose, a mathematical model of the radiation field inside a flat-plate heterogeneous reactor (a solar simulator) has been developed. The solid/liquid reactor is irradiated by two tubular UV lamps with the aid of two parabolic reflectors. Since titanium dioxide suspensions absorb and scatter radiation the model accounts for both efects. Resorting to information about the lamp and reflector characteristics, the catalyst optical properties and concentration, as well as the reactor dimensions and wall reactor properties, the solution of the mathematical model (a two-dimensional/two-directional model) provides a detailed description of the spatial and angular directional distributions of radiation intensities inside the reactor. Using this information, it is possible to precisely calculate the rate of absorbed radiation energy at each point inside the reactor. This is one of the key variables for reactor design and/or scale-up purposes. The radiation distribution inside the reactor was verified by computing forwardly transmitted and backwardly scattered radiation fluxes coming out of the reaction space through the glass reactor walls. These radiation fluxes were compared with experimental measurements made with a UV radiometer and good agreement was obtained; when no fouling was present and considering low catalyst loading to obtain measurable radiation fluxes, the maximum observed error was 12%. The predicted inlet boundary condition was also verified with actinometry and the error was smaller than 14% For this reactor configuration, when the radiation absorption performance is the only factor under consideration, it was found that Aldrich titania is more eficient than the Degussa P 25 variety.