In general, concentrations of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) measured in flue gas of municipal wastes incinerators (MWIs) vary significantly with waste component, combustion condition, operation stages and air pollution control devices equipped. Manual or short-time stack sampling process for dioxin-like congeners is the standard sampling method adopted by the Taiwan EPA (Method No. A807). However, it is difficult to include the dioxin emitted during unstable conditions, such as start-up process and abnormal condition. NCU-CS3 is a continuous stack sampling system developed by National Central University, Taiwan. In parallel with the Taiwan EPA standard sampling method (Method No. A807.75C), NCU-CS3 was used to monitor the emission of PCDD/Fs from a large-scale MWI in Kaohsiung city, Taiwan, including the periods of start-up and normal operation. Due to the unstable operating condition, the dioxin emission during start-up procedure (48 hours) could reach the peak of 50.8 ng I-TEQ/Nm3 before gradually decreasing in the following weeks. To reduce PCDD/F concentration in flue gas of the MWI, three measures have been carried out such as (1) reducing CO concentration, (2) removing the fly ash accumulated on the pipe and (3) alkali injection before the start-up procedure. The results obtained from the second measurements indicated the PCDD/Fs concentrations in flue gas measured by both NCU-CS3 and NIEA A807.75C were significantly lower than the first measurement and typically lower than 2.04 ng I-TEQ/Nm3. After 2 days’ operation, the PCDD/F emissions measured by NCU-CS3 and NIEA A807.75C were lower than the emission limit (0.1 ng I-TEQ/Nm3). In addition, the dioxin emissions of 168-hour continuous sampling with NCU-CS3 and NIEA A807.75C ranged from 0.072 to 0.753 ng I-TEQ/Nm3 (0.201 ng I-TEQ/Nm3 in average). Among these measurements (n=7), only three measurements are lower than 0.1 ng I-TEQ/Nm3. At the same time, 70 samples were collected with the NIEA A807.75C method (ten NIEA A807.75C samples correspond to each NCU-CS3 measurement). The measured dioxin emission was 0.0609 ng I-TEQ/Nm3 in average (0.0012~0.3835 ng I-TEQ/Nm3), while 77% samples were lower than the emission limit. The average dioxin emission factors measured by using NCU-CS3 and NIEA A807.75C methods were 0.845 and 0.346 μg I-TEQ/ton-waste, respectively. The effects of relevant operating parameters on PCDD/Fs emission are also evaluated via NCU-CS3, and they include: (1) alkali injection: HCl emission at the stack was controlled at 19 ppm (normal operation) and 15 ppm; (2) activated carbon injection: 0.24, 0.34 (normal operation), 0.44 and 0.55 kg / ton waste. Eighteen NCU-CS3 sampling results indicate that dioxins measured in smokestack decreased with the increase of AC injection rate. The dioxin removal efficiency achieved with ACI + BF technology was between 98.7% and 99.4%. Results indicate that dioxin concentration in 2nd cartridge of the NCU-CS3 increases with increasing cooling temperature. At 50oC, 5.4% of total dioxin concentration penetrated into the 2nd stage XAD-2 cartridge. As the cooling temperature was lower than 20oC, penetration was lower than 1% and could be ignored. The NCU-CS3 system showed reliable results with 70-130% recoveries of surrogate standard, which meets the regulation of standard method (NIEA A807.75C). This new system is capable of automatic, continuous sampling over a long period of time (≤4 weeks), and can accurately measure the PCDD/Fs emitted from MWIs. The PCDD/F emission from the MWI investigated included the start-up procedure via the NCU-CS3 system was around 150 mg I-TEQ/yr. The PCDD/F emission contributed by the first and second start-up procedures were 52% and 7%, respectively. Hence, the PCDD/F emission cannot be accurately quantified by relying on the manual and short-time stack sampling only. NCU-CS3 developed can provide more accurate data for government to enact effective control strategy.

polychlorinated dibenzo-p-dioxins, municipal wastes incinerators