Experimental and statistical determination of indicator parameters for the evaluation of fly ash and boiler ash...

On-line detectable indicator parameters in the flue gas of municipal solid waste incinerators (MSWI) such as chlorinated benzenes (PCBz) are well known surrogate compounds for gas-phase PCDD/PCDF concentration. In the here presented work derivation of indicators is broadened to the detection of fly and boiler ash fractions with increased PCDD/PCDF content. Subsequently these fractions could be subject to further treatment such as recirculation in the combustion chamber to destroy their PCDD/PCDF and other organic pollutants' content. Aim of this work was to detect suitable on-line detectable indicator parameters in the gas phase, which are well correlated to PCDD/PCDF concentration in the solid residues. For this, solid residues and gas-phase samples were taken at three MSWI plants in Bavaria. Analysis of the ash content from different plants yielded a broad variation range of PCDD/PCDF concentrations especially after disturbed combustion conditions. Even during normal operation conditions significantly increased PCDD/PCDF concentrations may occur after unanticipated disturbances. Statistical evaluation of gas phase and ash measurements was carried out by means of principal component analysis, uni- and multivariate correlation analysis. Surprisingly, well known indicators for gas-phase PCDD/PCDF concentration such as polychlorinated benzenes and phenols proved to be insufficiently correlated to PCDD/PCDF content of the solid residues. Moreover, no single parameter alone was found appropriate to describe the PCDD/PCDF content of fly and boiler ashes. On the other hand, multivariate fitting of three or four parameters yielded convenient correlation coefficients of at least r=0.8 for every investigated case. Thereby, comprehension of plant operation parameters such as temperatures and air flow alongside concentrations of inorganic compounds in the gas phase (HCl, CO, SO2, NOx) gave the best results. However, the suitable set of parameters suited best for estimation of PCDD/PCDF concentration in solid residues has to be derived anew for each individual plant and type of ash.