Despite the success of adsorption and thermal incineration of (C)VOC emissions, there is still a need for research on techniques which are both economically more favorable and actually destroy the pollutants rather than merely remove them for recycling elsewhere in the biosphere. The catalytic destruction of (C)VOC to CO2, H2O and HCl/Cl-2 appears very promising in this context and is the subject of the present paper. The experiments mainly investigate the catalytic combustion of eight target compounds, all of which are commonly encountered in (C)VOC emissions and/or act as precursors for the formation of PCDD/F. Available literature on the different catalysts active in the oxidation of (C)VOC is reviewed and the transition metal oxide complex V2O5-WO3/TiO2 appears most suitable for the current application. Different reactor geometries (e.g. fixed pellet beds, honeycombs, etc.) are also described. In this research a novel catalyst type is introduced, consisting of a V2O5-WO3/TiO2 coated metal fiber fleece. The conversion of (C)VOC by thermo-catalytic reactions is governed by both reaction kinetics and reaction equilibrium. Full conversion of all investigated VOC to CO2, Cl-2, HCl and H2O is thermodynamically feasible within the range of experimental conditions used in this work (260-340 degreesC, feed concentrations 30-60 ppm). A first-order rate equation is proposed for the (C)VOC oxidation reactions. The apparent rate constant is a combination of reaction kinetics and mass transfer effects. The oxidation efficiencies were measured with various (C)VOC in the temperature range of 260-340 degreesC. Literature data for oxidation reactions in fixed beds and honeycomb reactors are included in the assessment. Mass transfer resistances are calculated and are generally negligible for fleece reactors and fixed pellet beds, but can be of importance for honeycomb monoliths. The experimental investigations demonstrate: (i) that the conversion of the hydrocarbons is independent of the oxygen concentration, corresponding to a zero-order dependency of the reaction rate; (ii) that the conversion of the hydrocarbons is a first-order reaction in the (C)VOC; (iii) that the oxidation of the (C)VOC proceeds to a higher extent with increasing temperature, with multiple chlorine substitution enhancing the reactivity; (iv) that the reaction rate constant follows an Arrhenius dependency. The reaction rate constant k(r) (s(-1)) and the activation energy E (kJ/mol) are determined from the experimental results. The activation energy is related to the characteristics of the (C)VOC under scrutiny and correlated in terms of the molecular weight. The k(r)-values are system-dependent and hence limited in design application to the specific VOC-catalyst combination being studied. To achieve system-independency, k(r)-values are transformed into an alternative kinetic constant K (m(3)/(m(2) u)) expressed per unit of catalyst surface and thus independent of the amount of catalyst present in the reactor. Largely different experimental data can be fitted in terms of this approach. Results are thereafter used to define the Arrhenius pre-exponential factor A*, itself expressed in terms of the activation entropy. Destruction efficiencies for any given reactor set-up can be predicted from E- and A*-correlations. The excellent comparison of predicted and measured destruction efficiencies for a group of chlorinated aromatics stresses the validity of the design approach. Since laboratory-scale experiments using PCDD/F are impossible, pilot and full-scale tests of PCDD/F oxidation undertaken in Flemish MSWIs and obtained from literature are reported. From the data it is clear that: (i) destruction efficiencies are normally excellent; (ii) the efficiencies increase with increasing operating temperature; (iii) the higher degree of chlorination does not markedly affect the destruction efficiency. Finally, all experimental findings are used in design recommendations for the catalytic oxidation of (C)VOC and PCDD/F. Predicted values of the acceptable space velocity correspond with the cited industrial values, thus stressing the validity of the design strategy and equations developed in the present paper. (C) 2004 Elsevier B.V. All rights reserved.
