| dc.description.abstract | Around 60 to 80% of the seventeen 2,3,7,8-substituted PCDD/F concentrations in the atmosphere are bounded to particles. Partitioning of PCDD/F congeners between vapor and solid phases in flue gas of the PCDD/F emission sources and ambient air in Taiwan are evaluated via stack sampling and analysis in this study. This dissertation emphasizes the understanding of the partitioning and removal efficiency of PCDD/Fs of flue gas at two municipal wastes incinerator (MWI-1 and MWI-2), two industrial wastes incinerators (IWI-1 and IWI-2), one electric arc furnace (EAF) and one Waelz plant equipped with different types air pollution control devices (APCDs). The results indicate that the vapor-phase PCDD/Fs can be emitted from the stack by penetrating through cyclone (CY), bag filter (BF) and electrostatic precipitator (EP) if no effective control device is applied. Vapor-phase PCDD/Fs can be removed by various means including adsorption with carbon-based adsorbents, and catalytic destruction. Compared to the activated carbon injection technology which only transfers vapor-phase PCDD/Fs to the fly ash and would make ash disposal even more complicated, selective catalytic reduction (SCR) system can destroy PCDD/Fs and serves as a better control technology for removing PCDD/Fs from gas streams. The results of the emission from several facilities demonstrate that 99.7% and 0.3% of PCDD/F output in MWI-2 is discharged with EP ash (98.3 µg-TEQ/ ton waste) and stack gas, respectively. SCR system removes and destroys most of the PCDD/F congeners. The emission rate of MWI-1 is much higher than that of MWI-2 caused by the PCDD/F removal efficiencies achieved with different APCDs adopted, resulting in different PCDD/F removal mechanisms. It is noted that total PCDD/F discharge in Waelz plant is 840.3 µg-TEQ/ton EAF-dust, among which 33.3% is discharged with fly ash and needs to effectively reduce PCDD/F formation and install better PCDD/F control devices for the perspective of total environmental management. The results obtained from the ambient air sampling indicate that the mean PCDD/F concentration measured in the vicinity area of the MWI (56~348 fg-I-TEQ/m3) and EAF (61~312 fg-I-TEQ/m3) investigated are lower than the ambient air standard proposed in Japan (600 fg-I-TEQ/m3). The results obtained on vapor/solid partitioning of PCDD/Fs in ambient air indicate that the solid-phase portion accounts for more than 80% of the total concentration in the vicinity area of MWI investigated. Besides, the vapor-phase PCDD/Fs account for 35% to 55% in the vicinity area of EAF investigated. In addition, the temperature and the distance between emission source and sampling site would also affect the partitioning of PCDD/Fs between vapor and solid phases. The results of MWI-1 flue gas sampling indicate that there is optimal operating temperature for PCDD/F removal with ACI. In addition, the results of pilot-scale adsorption system (PAS) experimentation indicate that about 50% and 20% vapor-phase PCDD/Fs transferred to solid phase at Group 1 (1500C) and Group 2 (2000C), respectively. As the temperature is increased to 2500C, de novo synthesis significantly affects the partitioning of PCDD/Fs between vapor/solid phases.
Based on results of the partition of vapor/solid-phase PCDD/F achieved with the APCDs applied upstream and the particulate matter concentration in flue gas, this dissertation applies the equation log(Cv/Cs) =m logP0L+log(c/PM) for predicting vapor/solid-phase PCDD/F partition in flue gases downstream various APCDs including CY, EP, BF, wet electrostatic precipitator (WEP), fixed activated carbon bed (FCB), ACI and SCR. As the PM concentration is over 20,000 mg/Nm3 or temperature in flue gas is within the temperature window of de novo synthesis, the log(Cv/Cs) of observed data is significantly higher or lower than the result predicted, respectively. Accordingly, the equation can be used to predict the partitioning of PCDD/Fs between vapor and solid phases in flue gas if de novo synthesis is not significant. | en_US |