Article pubs.acs.org/est
Pilot Tests on the Catalytic Filtration of Dioxins Pao Chen Hung,† Shu Hao Chang,† Syuan Hong Lin,† Alfons Buekens,‡ and Moo Been Chang*,† †
Institute of Environmental Engineering, National Central University, 300 Jhong-da Road, Jhongli, Taoyuan 32001 Taiwan, Republic of China ‡ State Key Laboratory of Clean Energy Utilization, Zhejiang University, 38 Zheda Road, Hangzhou, Zhejiang 310027, People’s Republic of China ABSTRACT: Tests were conducted to study the removal efficiencies (REs) of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) from flue gas during a test program involving a pilot-scale catalytic filter (CF) module and a full-scale municipal solid waste incinerator (MSWI). The REs attained with the CF on a side stream and a conventional activated carbon (AC) injection and baghouse filtration system in the full-scale MSWI are evaluated via simultaneous sampling and analysis of both gas- and particle-phase PCDD/Fs. Flue gas without AC is supplied to the pilot-scale CF module for evaluating its RE capabilities. The REs achieved with the CF at 180 °C are 96.80 and 99.50%, respectively, for the gas phase and the particulate contained. The gas-phase PCDD/F RE rises significantly at 200 and 220 °C. The air/cloth (A/C) ratio defined as is the gas flow rate (m3/min) divided by the filtration area (m2) also affects the PCDD/F RE, especially in the gas phase. At 180 °C, a RE of gas-phase PCDD/Fs of 95.94% is attained with the CF at 0.8 m/min, yet it decreases at higher A/C ratios (1 and 1.2 m/min). A significantly lower toxic equivalency (TEQ) concentration (0.71 ng I-TEQ/ g) was measured in the filter dust of the CF module compared to that collected by the AC adsorption system (4.18 ng I-TEQ/g), apparently because of the destruction of gas-phase PCDD/Fs by the catalyst.
1. INTRODUCTION Incineration of municipal solid waste (MSW) results in the formation and partial emission of various semi-volatile products of incomplete combustion, including polychlorinated dibenzop-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). A total of 17 2,3,7,8-substituted PCDD/Fs are especially of concern, because of their high toxic equivalency (TEQ), bioaccumulation, and carcinogenicity.1,2 Incineration is a major contributor to global PCDD/F emissions, and these PCDD/Fs are also targeted by the Stockholm Convention, an international environmental treaty aiming to eliminate or restrict the production and use of persistent organic pollutants (POPs).3 Consequently, stringent regulations limiting dioxin emission concentrations (e.g., to 0.1 ng I-TEQ/Nm3) have been enforced in many countries.4 Solid residues from flue gas cleaning tend to exceed the regulatory limit of 1.0 ng I-TEQ/g, set by the Taiwanese Environmental Protection Administration (EPA). Currently, in Taiwan, 24 large-scale municipal solid waste incinerators (MSWIs) are in operation. Over 6 million tons of municipal wastes are incinerated, and over 0.3 million tons of filter dust [including neutralization salts and activated carbon (AC)] are generated annually. AC injection, followed by bag filtration (AC adsorption system), and selective catalytic reduction (SCR) have both been used to reduce emissions from MSWIs.5−7 For AC adsorption, PCDD/Fs can be effectively adsorbed by powder activated carbon (PAC) injected into flue gas, and then PCDD/F-containing PAC © 2014 American Chemical Society
and particulate matter are simultaneously separated by bag filtration. Literature indicates that the PCDD/F emission concentration with PAC injection is ca. 1% of that without PAC injection.8 However, AC just transfers PCDD/Fs from flue gas to solid residues. The SCR system is commonly applied to reduce NOx emissions from coal-fired power plants. Liljelind et al. indicate that the V2O5−WO3/TiO2 catalyst designed for the removal of nitrogen oxides (NOx) via SCR is also effective in decomposing PCDD/Fs at the same temperature used for NOx removal.9 Chang et al. also indicate that SCR of the metal smelting plant and MSWI in Taiwan can remove >60 and >95% of gas-phase PCDD/Fs, respectively.10 About 30% of the filter dust generated from MSWI in Taiwan contains more than 1.0 ng I-TEQ/g. Although SCR can actually destroy PCDD/Fs, it also shows some constraints, including high investment cost and deactivation of the catalyst, which limit its application to only a small fraction of MSWI plants. For example, in Taiwan, only one large-scale MSWI applies SCR for dioxin removal. A third option is catalytic filtration, combining the functions of filtration with those of catalytic destruction. A catalytic filter (CF) system consists of an expanded polytetrafluoroethylene (ePTFE) membrane and a catalytic felt substrate, which simultaneously remove solid-phase PCDD/Fs and destroy Received: Revised: Accepted: Published: 3995
November 6, 2013 February 25, 2014 February 27, 2014 February 27, 2014 dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
Article
Figure 1. Schematics of the (a) complete MSWI flowchart and (b) position of the pilot-scale CF module.
drying acid gas absorber and the bag filter (BF), and raw yet AC-free flue gas is available for testing and comparison. For the MSWI investigated in this study, a liquor of Ca(OH)2 and water mixed as the neutralizing agent is injected into the spray dry absorber with injection rates of 144 and 2000 kg/h, respectively, for removing acid gases, such as HCl and SO2. After injection of hydrated lime slurry and water and prior to AC injection, a small part of the flue gas was split off and introduced into the CF module. A 6 kg-AC/h injection rate is applied for AC adsorption, resulting in an AC concentration of 70 mg/Nm3 in the MSWI flue gas. The CF flue gas only contains particulate matter as well as some spent hydrated lime. To simulate the operating conditions of fabric filters in a real plant, a pilot-scale CF module with a CF (the CF module) is fabricated, as shown in Figure 1b. Real flue gas of MSWI is aspired into the CF module by an induced draft fan (IDF). Flue gas passes through the heating region, filter chamber, and IDF, and then it flows back to the exhaust duct of the MSWI. First, the flue gas temperature is raised to the designated operating temperature (180−220 °C) of the CF. Four CF sleeves (80 cm long and diameter of 15.5 cm) remove both the particulate matter (filtration) and the organic compounds (catalytic oxidation) from flue gas. The total filtration area is 1.63 m2. Two sets of air pulse jet periodically throw off the filter cake formed from the CF sleeves. Collected filter cake is stored for
gas-phase PCDD/Fs from the flue gas of thermal processes.11,12 Gas-phase PCDD/Fs are converted mainly into CO2, H2O, and HCl.13 Furthermore, the CF system also features a high removal efficiency (RE) for particulate matter.14 A pilot-scale module with CF bags (designated as the pilot-scale CF module) is installed in a large-scale MSWI located in eastern Taiwan. Real flue gas is split after the spray dry absorber but prior to AC injection, and a split stream is fed into the CF module for evaluating the collection, removal, and discharge characteristics of the 17 toxic PCDD/F congeners (Figure 1). This allows us to limit the impact of the considerable variation of raw flue gas conditions.
2. METHODS AND SAMPLE ANALYSIS This study aims to test the destruction of gaseous PCDD/Fs and the removal of solid-phase PCDD/Fs via catalytic filtration. For this purpose, a catalyst carrier (felt) is combined with V2O5−WO3/TiO2 catalytic powder. More details are described by Weber et al.15 Traditional AC adsorption has, however, been applied in most of the 24 large-scale MSWIs in Taiwan. To compare the PCDD/F destruction and removal efficiencies (DREs) achieved with a CF and an AC adsorption system, a large-scale MSWI located in eastern Taiwan (Figure 1a) was selected, because the injection point of PAC in this plant is set between the spray 3996
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
Article
3. RESULTS AND DISCUSSION 3.1. Characteristics of PCDD/Fs in Inlet Flue Gas. A PCDD/F-containing flue gas stream is supplied without any AC (yet with spent hydrated lime) from the MSWI investigated in this study (Figure 1). The O2 and CO2 contents of the gas stream are measured as 7−9.2 and 8.8−10.4 vol %, respectively. The relatively high water vapor content (≥25%) may be attributed to the high moisture content of the MSW incinerated and to a minor extent to the humid combustion air and the Ca(OH)2 slurry injected into the spray dry absorber. The particle concentrations in raw flue gas vary between 779 and 1350 mg/Nm3. This range is wide enough to influence the partition of PCDD/Fs between gas and particulate (G/P). Figure 2 shows the congener concentrations of the 17 2,3,7,8-
about 3 h in a dust hopper and then discharged into the collection tank (filter dust). Two temperature sensors are set at the inlet and outlet of the filter chamber, respectively, for making sure that the operating temperature is well-controlled. The CF used is similar to the GORE REMEDIA product tested by Bonte et al.11 To evaluate the REs of particulate matter and PCDD/Fs, two sampling points are set, at the inlet and outlet of the filter chamber, respectively. A Pitot tube is installed at the outlet of the filter chamber to monitor the gas flow rate, and the air/ cloth ratio [A/C ratio (m/min) = gas flow rate (m3/min)/ filtration area (m2)] can be further controlled by establishing different operating conditions. In this study, the effect of the A/ C ratio (varying from 0.8 to 1.2 m/min) on the PCDD/F RE will be investigated. Its impact is directly comparable to that of the gaseous hourly space velocity (h−1), except for its dimension. Before conducting flue gas PCDD/F sampling, the operating conditions of the pilot-scale CF module were adjusted and stabilized for 1 day. The sampling procedure of flue gas follows isokinetic sampling of the United States Environmental Protection Agency (U.S. EPA) (Method 5). In the U.S. EPA Method 23, the solid- and gas-phase PCDD/Fs are simultaneously collected with a glass fiber filter and XAD-2 resin, respectively. A total of 2−3 h is needed for each PCDD/ F sampling. Moreover, PCDD/F samples are simultaneously collected at the inlet and outlet of CF or AC adsorption systems. The results of PCDD/F sampling are further used to calculate the REs achieved with the CF and AC adsorption, respectively. The samples of flue gas (including solid- and gasphase samples) and the filter dust collected were spiked with known amounts of Method 23 internal standard and Method 1613 labeled standards, following quantification standards, respectively, and then analyzed for the 17 2,3,7,8-substituted PCDD/F congeners with high-resolution gas chromatography (HRGC, Thermo Trace GC)/high-resolution mass spectrometer (HRMS, Thermo DFS) using a fused silica capillary column DB-5 MS (60 m × 0.25 mm × 0.25 μm). In this study, all TEQ concentrations of PCDD/Fs are calculated with international toxic equivalent factors (I-TEFs). The REs of pollutants in flue gas and filter dust are both calculated with the following formula: RE (%) =
Figure 2. Profiles of the 17 toxic PCDD/F congeners in the flue gas at the inlet of the catalytic test unit: (a) mass concentration and (b) TEQ concentration.
substituted PCDD/Fs at the inlet sampling point of the CF module. In the PCDD/F congener profile, three major congeners appear: OCDD (3.41 ng/Nm3 in the gas phase and 6.08 ng/Nm3 in the particulate contained), 1,2,3,4,6,7,8HpCDF (5.60 ng/Nm3 in the gas phase and 5.22 ng/Nm3 in the particulate contained), and OCDF (4.68 ng/Nm3 in the gas phase and 5.79 ng/Nm3 in the particulate contained). The PCDD/F mass concentration and the TEQ concentration of the raw flue gas are 53.5 ng/Nm3 and 2.87 ng I-TEQ/Nm3, respectively. Figure 2 also indicates that the highly chlorinated congeners dominate the PCDD/Fs present in the particulate matter of inlet flue gas when expressed in weight percent (Figure 2a), while the low chlorinated congeners control the PCDD/Fs when expressed in TEQ units (Figure 2b). The difference is caused by the vapor pressure of PCDD/F congeners and even more by the effect of the TEQ factors. The results are similar to those reported by other studies.16−18 As usual, 2,3,4,7,8-PeCDF contributes most to the TEQ concentration profile. The TEQ concentration profile also indicates that gas-phase PCDD/Fs account for more than 60% of the total TEQ.
(concentration)inlet − (concentration)outlet (concentration)inlet
which ignores potential dilution by air entering into the flue gas stream (mainly in the MSWI), as well as possible phase transfer, i.e., adsorption of PCDD/Fs from the gas onto filter cake (gas to particle) and desorption of PCDD/Fs from dust to gas in clean gas (particle to gas). Furthermore, discharge factors, including both the stack emission and discharge from air pollution control devices of AC adsorption and catalytic filtration, are discussed in this study for understanding the environmental impact achieved with different control technologies and defined as below discharge factor (μg I‐TEQ/ton MSW) = (flue gas flow rate × PCDD/F concentration in flue gas + fly ash discharge rate × PCDD/F concentration in fly ash)/mass of MSW incinerated 3997
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
Article
removal by the CF. Figure 3a indicates that the REs of PCDD congeners are slightly lower than those of PCDF congeners. A similar trend was also observed when the SCR catalyst (V2O5−WO3/TiO2) was applied.15 Debecker et al. investigated the effectiveness of three different catalysts for PCDD/F destruction, including V2O5/TiO2−SO4, V2O5−MoO3/TiO2, and V2O5−WO3/TiO2.19 However, the trends of destruction between PCDDs and PCDFs are different. For V2O5/TiO2− SO4 and V2O5−MoO3/TiO2, the destruction efficiency of PCDFs is higher than that of PCDDs. However, the destruction efficiency of PCDFs is lower than that of PCDDs as V2O5− WO3/TiO2 is applied. Smolka and Schmidt indicated that low-volatile congeners were adsorbed comparatively stronger than high-volatile congeners.20 Higher chlorinated PCDD/F congeners are of lower vapor pressures compared to low-chlorinated PCDD/Fs and, thus, more easily adsorb onto particulate matter. Because the operating temperature is lower than the boiling point of PCDD/F congeners, PCDD/F congeners may effectively adsorb onto particulate matter and further be significantly removed by membrane filtration of the CF. Figure 3 also shows the efficiencies of AC adsorption for PCDD/F removal in the full-scale MSWI investigated, in parallel with the CF. REs of PCDD/Fs in the gas phase and particulate matter contained are 97.68 and 98.65%, respectively. These PCDD/F REs are achieved with the AC adsorption system operating at 150 °C. The adsorption efficiency varies with several parameters, such as the operating temperature, amount of AC injected, amount of particulate matter in the flue gas, etc.21 The results show that AC adsorption can effectively adsorb the gas-phase PCDD/F congeners and collect the solidphase PCDD/Fs. REs of solid-phase PCDD/Fs of all congeners achieved with AC adsorption are relatively constant at a given temperature. With regard to PCDD/F congeners, Mori et al. indicate that highly Cl-substituted PCDD/F congeners can be effectively adsorbed by AC because they have relatively high boiling points.22 Furthermore, reduction of gas-phase PCDD/F REs with an increasing Cl-substituted level via the AC adsorption system investigated is also observed in this study, especially REs of PCDD isomers. In contrast, less chlorinated congeners of PCDD/Fs are more volatile than highly chlorinated congeners; they are better represented in the gas phase and, hence, better adsorbed by AC. Adsorption and filtration are the two main mechanisms of AC adsorption, and as expected, the volatility of the PCDD/F congeners was the most important factor influencing PCDD/F REs. To understand the influence of the catalytic contact time, three A/C ratios (0.8, 1.0, and 1.2 m/min) were selected for operating the CF module. The operating temperature of 200 °C is designated as the baseline condition. Fritsky et al. indicate that gas-phase PCDD/Fs in the CF system must go through one or all of the following steps: (1) adsorbing on filter media, (2) going through these filter media and eventually being emitted, (3) adsorbing on particulate matter, or (4) being destroyed by the catalyst.5 However, this study indicates that most of the gas-phase PCDD/Fs are destroyed by the catalyst. These results are consistent with the laboratory-scale study conducted by Weber et al.23 Figure 4 shows the REs of toxic PCDD/Fs, including the gas/solid phase, total toxic PCDD/Fs, and particulate matter, with different A/C ratios. The REs of PCDD/Fs on particulate achieved with the CF at 0.8, 1.0, and 1.2 m/min are 99.74, 99.47, and 99.63%, respectively. The REs of particulate matter achieved with the CF operating at an A/C
The average concentration of particulate matter measured in raw flue gas (yet after injection of hydrated lime) is 1120 mg/ Nm3, while the average mass concentration of solid-phase PCDD/Fs is 23.7 ng/Nm3, suggesting a unit load of the order of 21 ng/g or 0.76 ng I-TEQ/g. Solid-phase PCDD/F is significantly higher for gas streams containing a high concentration of particulate matter. Even so, gas-phase PCDD/F dominates the total TEQ concentration. Average TEQ concentrations of solid- and gas-phase PCDD/Fs are 0.95 ± 0.35 and 1.92 ± 0.07 ng I-TEQ/Nm3, respectively (4 sampling times). 3.2. PCDD/F Removal Achieved with the CF and AC Adsorption. Three temperatures (180, 200, and 220 °C) are selected to evaluate the temperature effect on the PCDD/F RE, achieved with the CF module. In general, catalyst activity increases with a rising temperature. On the other hand, a rising gas temperature would raise the volatility of the PCDD/F congeners present in filter cake. Therefore, overall REs may also be reduced at enhanced temperatures if higher catalyst activity does not compensate for the effects of rising volatility and declining catalytic adsorption. Figure 3 shows the REs of PCDD/F congeners in the gas/ solid phases attained with the CF module at three different
Figure 3. REs of PCDD/F congeners achieved with the pilot-scale CF module at different operating temperatures and the AC adsorption system as adopted in the MSWI: (a) gas-phase PCDD/F congeners and (b) solid-phase PCDD/F congeners.
operating temperatures. The total REs of PCDD/F in the gas phase achieved with the CF at 180, 200, and 220 °C are 96.80, 98.42 and 99.04%, respectively. Therefore, catalytic PCDD/F destruction activity rises significantly at a higher operating temperature, even though more PCDD/Fs are vaporized because of the increase of the temperature. On the other hand, the REs of solid-phase PCDD/Fs achieved with the CF remain almost constant, with 99.50, 99.63 and 99.72% for 180, 200, and 220 °C, respectively. These values are lower than those for the particulate collection efficiency, i.e., 99.80, 99.83, and 99.82 wt % for 180, 200, and 220 °C, respectively. Overall, increasing the temperature is beneficial for enhancing PCDD/F 3998
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
Article
Figure 4. REs achieved with the pilot-scale CF module operated at different A/C ratios (operating temperature of 200 °C): (a) gas-phase toxic PCDD/Fs, (b) solid-phase toxic PCDD/Fs, (c) total toxic PCDD/Fs (including gas- and solid-phase toxic PCDD/Fs), and (d) particulate matter.
ratio of 0.8, 1.0, and 1.2 m/min are 99.79, 99.87, and 99.81%, respectively. The main mechanism leading to the removal of particulate matter and solid-phase PCDD/Fs is clearly by filtration. Therefore, the effect of A/C ratios on the REs of particulate matter and solid-phase PCDD/Fs remains slight. On the other hand, the REs of gas-phase PCDD/Fs achieved with the CF at 0.8, 1.0, and 1.2 m/min are 98.75, 98.50, and 98.31%, respectively. A previous study indicates that the PCDD/F RE achieved with SCR decreases with the rising space velocity.7 For gas-phase PCDD/Fs, the TEQ RE of gas-phase PCDD/Fs slightly reduces from 98.75 to 98.31% as the A/C ratio is increased from 0.8 to 1.2 m/min. This RE compares poorly with that of particulate matter and solid-phase PCDD/Fs. The A/C ratio mainly affects the residence time of pollutants within the catalytic felt substrate. Within the average depth of catalytic felt substrate (1.5 mm), the residence times are only 0.113, 0.090, and 0.075 s, with an A/C ratio of 0.8, 1.0, and 1.2 m/ min, respectively. Obviously, the gas residence time has a great impact on the REs of gas-phase PCDD/Fs. Therefore, these results clearly indicate that the PCDD/F REs achieved with the pilot-scale CF module operating with an A/C ratio of 1 m/min at 180 °C are close to those achieved by AC adsorption operating with the same A/C ratio at 150 °C. However, the PCDD/F REs achieved with the CF are still significantly enhanced with an increasing temperature (200 and 220 °C). On the other hand, the PCDD/F REs achieved with the CF vary significantly with the A/C ratio. On the basis of the different removal mechanisms, increasing REs are found with AC adsorption with more Cl-substituted PCDD/F congeners in flue gas. 3.3. PCDD/F Congeners in Filter Dust Discharged from the CF Module and AC Adsorption System. Both the CF and AC adsorption proved effective in removing gasphase PCDD/Fs. However, a significant difference of the PCDD/F content in filter dust is demonstrated in this study between the CF dust from the CF module and the BF dust from an AC adsorption system. For purposes of comparison, the PCDD/F concentration in the CF dust is the PCDD/F contents with the CF operating at 180 °C, while the operating temperature of the BF in the MSWI investigated is maintained at 150 °C.
Figure 5 shows the concentrations of PCDD/Fs in the CF dust and BF dust. On the mass concentration profile, the most
Figure 5. Concentrations of PCDD/F congeners in flue gas cleaning residues from AC adsorption and CF technology (PCDD/F concentrations in the CF dust collected with 180 °C): (a) mass concentrations and (b) TEQ concentrations.
abundant congener is OCDD (3.49 ng/g in CF dust and 47.7 ng/g in BF dust) in both the CF dust and BF dust. The PCDD/F congener profile is similar to the characteristics of inlet flue gas. However, in the I-TEQ concentration profile, the major contributor is 2,3,4,7,8-PeCDF (0.22 ng I-TEQ/g in the CF dust and 1.33 ng I-TEQ/g in the BF dust). All I-TEQ concentrations of PCDD/F congeners in the CF dust are commonly lower than those in the BF dust. It is worthwhile to note that PCDD/F concentrations are 0.71 and 4.18 ng ITEQ/g in the CF dust and BF dust, respectively. The PCDD/F concentration in the BF dust is nearly 6 times the concentration in the CF dust and exceeds the limit set by the Taiwanese EPA 3999
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
Article
(1.0 ng I-TEQ/g). Therefore, higher PCDD/F concentrations are found on filter dust from the AC adsorption system compared to that of CF technology. Because of the good combination of the ePTFE membrane and the composition of catalyst, the CF provides two important functions, including catalytic destruction of gas-phase PCDD/Fs and filtration of particulate matter, including solid-phase PCDD/Fs. Besides, the catalyst could significantly reduce the concentrations of PCDD/Fs in the filter dust and that of gas-phase PCDD/Fs in flue gas discharged. Interestingly, no significant difference is found for PCDD/F concentrations in the CF dust when the CF was operated at a temperature range of 180−220 °C. This result indicates that the PCDD/F content in filter dust can be effectively reduced by applying catalytic filtration. On the other hand, a significant difference between low and highly Clsubstituted congeners is found in the BF dust but not for the CF dust; the adsorption achieved with AC is significantly enhanced by a higher chlorination level because of the high boiling points and low vapor pressures of high Cl-substituted PCDD/F congeners. 3.4. Comparing the CF to AC Adsorption. Additionally, the REs of PCDD/Fs achieved with the CF system tested and the AC adsorption system applied in the MSWI plant are further compared and discussed. Adsorption is the main mechanism responsible for removing gaseous organic pollutants by AC adsorption. Gaseous pollutants are effectively adsorbed by AC because of its extremely high specific surface area. However, all solid- and gas-phase pollutants are concentrated onto the AC, hydrated lime, and particulate matter. In contrast, with the CF, part of the gaseous organic compounds is adsorbed as it passes through the filter cake and the other part is largely destroyed with catalysts. In this study, relatively high REs of PCDD/Fs from flue gas are found for both the CF and AC adsorption but significantly higher PCDD/F concentrations remain in the BF dust compared to that in the CF dust. For a better understanding of the overall environmental impact presented by both technologies, discharge factors are calculated for both the CF and AC adsorption. Figure 6 shows the toxicity (TEQ) flow rates of PCDD/Fs via the two discharging pathways (stack emission and filter dust discharge) with AC adsorption and CF systems, respectively. The percentages of TEQ flow are derived by comparison to the average inlet TEQ flow rate of PCDD/Fs. The average input rate of PCDD/F TEQ is 30.74 μg I-TEQ/ton in both AC adsorption and CF systems. The emission rates of PCDD/Fs attained with AC adsorption and CF technology are 0.39 and 0.31 μg I-TEQ/ton, respectively, or 1.27 and 1.01% of the input rate, respectively. Thus, the TEQ stack emission rates achieved with both technologies are close. With AC adsorption, Figure 6a shows a total PCDD/F TEQ flow rate of 106% of the input TEQ flow rate. The extra 6% is presumably due to the closure error, even though, in principle, PCDD/Fs might be formed significantly because the AC injected can supply additional reaction area and carbon source. On the other hand, the operating temperature of 150 °C is normally too low to sustain PCDD/F formation. Regardless, Figure 6b clearly shows that the PCDD/F TEQ flow rate via filter dust discharge of CF technology is only 33.4% of the input TEQ flow rate, much lower than that for AC adsorption. For both technologies, the main discharging pathway is via filter dust. The output flow rate of PCDD/F TEQ via filter dust discharged with CF technology is only one-third of the output flow with an AC adsorption system. Hence, in comparison to AC adsorption, the CF is
Figure 6. Mass balance (TEQ flows): (a) AC adsorption system with 180 °C and (b) catalytic filtration system with 200 °C.
more environmentally friendly from the perspective of total PCDD/F emission. Finally, these results obtained in this study clearly indicate that catalytic filtration is a superior solution from the perspective of total environmental management and, thus, remains an air pollution control technology, which is worthy of being promoted. In contrast, PCDD/F contents in filter dust from the AC adsorption system are commonly higher than the PCDD/F limit of solid waste in Taiwan (1.0 ng I-TEQ/g). In addition, the CF is static during the operation, and gas-phase PCDD/Fs are effectively and continuously destroyed at a stable operating temperature. In comparison to catalytic filtration, the AC adsorption system is a dynamic system and easily affected with AC quality and AC bridging in the channel. However, further studies are needed to enhance application of catalytic filtration, such as lifetime of the CF applied in this study (longterm tests) and feasibility for multi-pollutant removal (NOx reduction). In later work of this study, the effect on other POPs, such as polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs) will be tested.
■
AUTHOR INFORMATION
Corresponding Author
*Telephone/Fax: +886-3-4226774. E-mail: mbchang@ncuen. ncu.edu.tw. Notes
The authors declare no competing financial interest. 4000
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
Environmental Science & Technology
■
Article
waste incinerator. J. Hazard. Mater. 2008, 160, 37−44, DOI: 10.1016/ j.jhazmat.2008.02.077. (17) Domingo, J. L.; Granero, S.; Schuhmacher, M. Congener profiles of PCDD/Fs in soil and vegetation samples collected near to a municipal waste incinerator. Chemosphere 2001, 40, 517−524, DOI: 10.1016/S0045-6535(00)00403-3. (18) Chang, M. B.; Chi, K. H.; Chang, S. H.; Chen, Y. W. Measurement of PCDD/F congener distributions in MWI stack gas and ambient air in northern Taiwan. Atmos. Environ. 2004, 38, 2535− 2544, DOI: 10.1016/j.atmosenv.2004.01.037. (19) Debecker, D. P.; Delaigle, R.; Hung, P. C.; Buekens, A.; Gaigneaux, E. M.; Chang, M. B. Evaluation of PCDD/F oxidation catalysts: Confronting studies on model molecules with tests on PCDD/F-containing gas stream. Chemosphere 2011, 82, 1337−1342, DOI: 10.1016/j.chemosphere.2010.12.007. (20) Smolka, A.; Schmidt, K. G. Gas/particle partitioning before and after gas purification by an activated carbon filter. Chemosphere 1997, 34, 1075−1082, DOI: 10.1016/S0045-6535(96)00409-2. (21) Lu, S. Y.; Ji, Y.; Buekens, A.; Ma, Z. Y.; Jin, Y. Q.; Li, X. D.; Yan, J. H. Activated carbon treatment of municipal solid waste incineration flue gas. Waste Manage. Res. 2013, 31, 169−177, DOI: 10.1177/ 0734242X12462282. (22) Mori, K.; Matsui, H.; Yamaguchi, N.; Nakagawa, Y. Multicomponent behavior of fixed-bed adsorption of dioxins by activated carbon fiber. Chemosphere 2005, 61, 941−946, DOI: 10.1016/ j.chemosphere.2005.03.035. (23) Weber, R.; Plinke, M.; Xu, Z. Dioxin destruction efficiency of catalytic filters evaluation in laboratory and comparison to field operation. Org. Compd. 2000, 45, 427−430.
ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support provided by the National Science Council (NSC 100-2221-E008-017-MY3) and EPA of the Republic of China.
■
REFERENCES
(1) Behnisch, P. A.; Hosoe, K.; Sakai, S. Combinatorial bio/chemical analysis of dioxin and dioxin-like compounds in waste recycling, feed/ food, humans/wildlife and the environment. Environ. Int. 2001, 27, 495−519, DOI: 10.1016/S0160-4120(01)00029-0. (2) Sakai, S.; Deguchi, S.; Takatsuki, H.; Uchibo, A. Large-scale fires and time trends of PCDDs/DFs in sediments. Chemosphere 2001, 43, 537−547, DOI: 10.1016/S0045-6535(00)00405-7. (3) Jones, P. H.; de Gerlache, J.; Marti, E.; Mischer, G.; Scherrer, M. C.; Bontinck, W. J.; Niessen, H. J. The global exposure of man to dioxins: A perspective on industrial waste incineration. Chemosphere 1993, 26, 1491−1497, DOI: 10.1016/0045-6535(93)90216-R. (4) Hai, J.; Zhang, G.; Cheng, J. Characterization and mass balance of dioxin from a large-scale municipal solid waste incinerator in China. Waste Manage. 2012, 32, 1156−1162, DOI: 10.1016/j.wasman.2012.01.024. (5) Chang, M. B.; Lin, J. J. Memory effect on the dioxin emissions from municipal waste incinerator in Taiwan. Chemosphere 2001, 45, 1151−1157, DOI: 10.1016/S0045-6535(00)00571-3. (6) Chang, M. B.; Yang, C. C.; Chang, S. H.; Hong, B. Z.; Chi, K. H. Innovative PCDD/F-containing gas stream generating system applied in catalytic decomposition of gaseous dioxins over V2O5−WO3/TiO2based catalysts. Chemosphere 2008, 73, 890−895, DOI: 10.1016/ j.chemosphere.2008.07.027. (7) Liljelind, P.; Unsworth, J.; Maaskant, O.; Marklund, S. Removal of dioxins and related aromatic hydrocarbons from flue gas streams by adsorption and catalytic destruction. Chemosphere 2001, 42, 615−623, DOI: 10.1016/S0045-6535(00)00235-6. (8) Chi, K. H.; Chang, S. H.; Chang, M. B. PCDD/F emissions and distributions in Waelz plant and ambient air during different operating stages. Environ. Sci. Technol. 2007, 41, 2515−2522, DOI: 10.1021/ es061665p. (9) Liljelind, P.; Unsworth, J.; Maaskant, O.; Marklund, S. Removal of PCDD/Fs and related aromatic hydrocarbons from flue gas streams by adsorption and catalytic destruction. Chemosphere 2001, 42, 615− 23, DOI: 10.1016/S0045-6535(00)00235-6. (10) Chang, M. B.; Chi, K. H.; Chang, S. H.; Yeh, J. W. Destruction of PCDD/Fs by SCR from flue gases of municipal waste incinerator and metal smelting plant. Chemosphere 2007, 66, 1114−1122, DOI: 10.1016/j.chemosphere.2006.06.020. (11) Bonte, J. L.; Fritsky, K. J.; Plinke, M. A.; Wilken, M. Catalytic destruction of PCDD/F in a fabric filter: Experience at a municipal waste incinerator in Belgium. Waste Manage. 2002, 22, 421−426, DOI: 10.1016/S0956-053X(02)00025-9. (12) Šyc, M.; Pekárek, V.; Fišerová, E.; Punčochár,̌ M.; Karban, J.; Prokeš, O. Catalytic filter application in the TERMIZO municipal waste incineration plant in Liberec. Org. Compd. 2006, 68, 1232−1235. (13) Xu, Z.; Fritsky, K. J.; Graham, J.; Dellinger, B. Catalytic destruction of PCDD/F: Laboratory test and performance in a medical waste incinerator. Org. Compd. 2000, 45, 419−422. (14) Fritsky, K. J.; Kumm, J. H.; Wilken, M. Combine PCDD/F destruction and particulate control in a baghouse: Experience with a catalytic filter system at a medical waste incineration plant. J. Air Waste Manage. Assoc. 2001, 51, 1642−1649, DOI: 10.1080/ 10473289.2001.10464391. (15) Weber, R.; Plinke, M.; Xu, Z.; Wilken, M. Destruction efficiency of catalytic filters for polychlorinated dibenzo-p-dioxin and dibenzofurans in laboratory test and field operationInsight into destruction and adsorption behavior of semivolatile compounds. Appl. Catal., B 2001, 31, 195−207, DOI: 10.1016/S0926-3373(00)00278-2. (16) Lin, C.; Chen, C. K.; Lin, Y. C.; Wang, L. C.; Chang-Chien, G. P. Polychlorinated dibenzo-p-dioxins/dibenzofuran mass distribution in both start-up and normal condition in the whole municipal solid 4001
dx.doi.org/10.1021/es404926g | Environ. Sci. Technol. 2014, 48, 3995−4001
