Environ. Sci. Technol. 2008, 42, 6211–6217
Flow Analysis of Heavy Metals in MSW Incinerators for Investigating Contamination of Hazardous Components HUA ZHANG, PIN-JING HE,* AND LI-MING SHAO State Key Laboratory of Pollution Control and Resources Reuse, Key Laboratory of Yangtze River Water Environment, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, P.R. China
Received February 23, 2008. Revised manuscript received May 28, 2008. Accepted May 30, 2008.
The contribution of hazardous components in municipal solid waste (MSW) to environmental risks has seldom been quantified due to their heterogeneous streams and irregular disposal patterns. A material flow analysis, in which the input metals in major MSW compositions (excluding discriminable hazardous components) were subtracted from the total output metals in the treatment products, was proposed to estimate the heavy metal contamination in MSW due to hazardous components. The statistical data from 1-year field measurements for two largescale incinerators in Shanghai city were used as an illustrative example. The results indicated that the amount of Cr and Ni in the incineration products were similar to those found in the major MSW compositions, while the amounts of Cd, Cu, Pb, and Zn in the incineration products were 2.27-4.00 times, 1.90-3.77 times, 2.25-3.51 times, and 2.98-4.06 times greater than that in the MSW. According to evaluation, more than 56-75% of Cd, 47-74% of Cu, 56-72% of Pb, and 66-75% of Zn in the MSW were contributed by the minor hazardous components, indicating the need for source separation. The methodology provides a cost-effective procedure for quantification of the hazardous waste contamination in MSW.
Introduction Municipal solid waste (MSW) can be reused as an organic fertilizer or for soil amendment (1, 2) following certain transformation processes or to recover energy from its incineration (3, 4). However, the heavy metals contained in MSW and its transformation products make MSW hazardous to the environment, and this has led to increasing concerns as far as MSW management is concerned (5, 6). Source analysis of heavy metals in MSW compositions could allow policy makers and management authorities to effectively control the major contaminant streams and conduct integrated MSW management practices (7). The direct sampling and analysis of MSW compositions is the most common practice (8–10). Due to the great heterogeneity of MSW, a sufficient number of MSW samples should be provided in order to clearly demonstrate the heavy metal characteristics in MSW. By using material flow analysis (MFA) sa systematic assessment of the metal flows within a system * Corresponding author phone: +86-21-6598 6104; fax: +86-216598 6104; e-mail:
[email protected]. 10.1021/es800548w CCC: $40.75
Published on Web 07/02/2008
2008 American Chemical Society
defined in space and time, heavy metal contamination in MSW was discussed (5, 7, 8). The basis of the assessment is mass conservation. The MFA method is attractive as a decision-support tool in the evaluation of waste management options, allowing for an effective cause-effect modeling when associated with specific environmental effects. Hazardous components (e.g., used batteries, waste electrical and electronic equipment, garden chemicals, etc.) constitute a small fraction of MSW; however, some of these substances contain a high level of heavy metals (11, 12), which may cause problems. In many countries, separate collection of hazardous components is rare or not well established; consequently, these hazardous substances are disposed of along with nonhazardous waste (13). There have been some attempts to quantify hazardous MSW components by using questionnaire surveys (14–16), spot checks (17, 18), or product flow analysis (19). All this research has focused on the sources, types, and quantities of hazardous components or the characteristics of the hazardous products possibly discarded into MSW. Only a few studies have reported the percentage of heavy metals in MSW that originate from hazardous components, by estimating the rate at which products consumed end up as MSW (20, 21) or by surveying the quantity of hazardous components in MSW (5, 22, 23), on the basis of the heavy metal contents in the hazardous components. However, the number of households surveyed was usually as low as a few hundred, owing to the expensive, time-consuming nature of field assessments. It is more difficult to estimate the contamination of hazardous components in MSW than of the other major MSW compositions, due to the small amount and diverse types of hazardous wastes present in MSW, their irregular disposal patterns, and discrepancies in classification systems (17). Therefore, the evaluation of such a heterogeneous stream is rather difficult, unless a wider and long-term survey of the samples is conducted. By considering that the major MSW compositions and the resulting treatment products are more stable and that their characteristics are relatively more homogeneous and well-defined (17, 24), an MFA method in which the input metals in the major MSW compositions (excluding discriminable hazardous components) are subtracted from the total output metals in the treatment products, which has hitherto not been applied to hazardous MSW, is proposed in this study to demonstrate the heavy metal contamination caused by hazardous components. Procedures are presented to (a) analyze the heavy metal contents in the major MSW compositions (excluding discriminable hazardous components) and other inputs as well as the outputs from treatment facilities routinely with adequate frequency and (b) estimate the contribution of hazardous components to the heavy metals in the total MSW by subtracting the inputs (excluding hazardous MSW) from the outputs based on statistical analysis. One-year field measurements for two large-scale incinerators (serving more than 2 million people) in Shanghai city were conducted to guarantee the involvement of a sufficient number of households. The objective of this study was to present a new costeffective method for a more accurate quantification of the contamination of MSW due to hazardous streams and the evaluation of the effect of MSW management practices, by replacing the sampling of the minor hazardous components with the major MSW compositions and treatment products. More stable production, physical composition, and characteristics for the latter could effectively reduce the uncertainty during sampling and measurements. Furthermore, the VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6211
FIGURE 1. Material flow in (a) the MSW treatment facility and (b) the MSW incinerator. field assessment conducted in the treatment facilities, which receive MSW from most of the serving areas, could cover nearly all the waste from the producers (hardly realizable by spot check). Therefore, the time span of the research could be shortened (only periodical sampling and analysis is required), and only one facility was involved, instead of hundreds or thousands of households.
Materials and Methods Setup of MFA for Hazardous Components. Figure 1a shows the material flow in a MSW treatment facility. Based on the law of mass conservation, the flow balance can be described by eqs 1 and 2 (wet basis)
∑M i
∑
x Finputi )
i
∑ i
inputi )
∑M
(1)
outputj
j
x (Minputi × Cinputi ))
∑
x (Moutputj × Coutputj ))
j
∑F
x outputj
(2)
j
where Minput i and Moutput j ) the mass of input i and output j; Fxinput i and Fxoutput j ) the amount of heavy metal x contained x x in input i and output j; and Cinput i and Coutput j ) the concentration of heavy metal x in input i and output j. Figure 1b shows the material flow in an incinerator. Because of the negligible amount of heavy metals contained in the air to furnace and reaction reagent to air pollution control (APC) system (25), the heavy metal balance can be described by applying eq 3. Then the contamination of x ) and their contribution to hazardous components (FH-MSW x x + /(FBA the heavy metals in the total MSW [FH-MSW x +Fx x FAR leachate+Fgas)] can be deduced x x x x x x x + FH-MSW ) MMSW × CMSW + MH-MSW × CH-MSW ) FBA + FMSW x x x x x x x + Fleachate + Fgas ) MBA × CBA + MAR × CAR + FAR x x x x × Cleachate + Mgas × Cgas (3) Mleachate x x x , Fx , F x x , FH-MSW , FBA where FMSW AR leachate, and Fgas ) the total amount of heavy metal x contained in the incinerated nonhazardous MSW compositions, hazardous components, bottom ash, APC residues, leachate, and exhaust gas, x x x , Mx , M x x , MH-MSW , MBA respectively; MMSW AR leachate, and Mgas ) the mass of the incinerated nonhazardous MSW compositions, hazardous components, bottom ash, APC residues, x x , CH-MSW , leachate, and exhaust gas, respectively; and CMSW x , Cx , C x x CBA AR leachate, and Cgas ) the concentration of heavy metal x in the incinerated nonhazardous MSW compositions,
6212
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008
hazardous components, bottom ash, APC residues, leachate, and exhaust gas, respectively. The amount of heavy metal x contained in 1 t (wet) of x ) was calculated from eq 4. major MSW compositions (FMSW x ), bottom Based on the production ratio of leachate (Mleachate ash (MxBA), and APC residues (MxAR), the amount of heavy metal x , exhaust gas was contained in the incineration products (FIP not counted) on incinerating 1 t of wet MSW was converted as shown in eq 5 x x (g ⁄ t) ) CMSW × (1 - Wy) FMSW
(4)
x x x FIP (g ⁄ t) ) CAR × MAR × (1 - WAR) + CBA × MBA × (1 - WBA) + x × Mleachate (5) Cleachate
where WAR and WBA ) the water content of the APC residues and bottom ash, respectively. According to Belevi and Moench (26), less than 1% of heavy metals in the waste was emitted into the air through stack gas. Therefore, in view of the difficulty involved in monitoring, heavy metal emission with exhaust gas was not measured in this study. Sample Collection. The MSW samples and incineration residues were collected monthly from the two large-scale MSW incinerators (1200 t/d and 1500 t/d) in Shanghai during October 2004 to September 2005. The sampling procedures were presented in detail in the Supporting Information. In brief, 100-150 kg of the mixed MSW (within one day of production) was collected from 5-7 randomly selected locations in the unloading area of the MSW pits and then separated into nine major fractions. Distinguishable hazardous wastes such as batteries and electronic waste were removed from the sample. The heavy metal content in the major MSW compositions was calculated by summing the amount of the metal in each fraction by the dry weight. Bottom ash (100-150 kg) was collected from 10 randomly selected locations in the ash pits. After mixing, about onequarter was divided and ground into particles of sizes less than 154 µm before heavy metal analysis. Approximately 20 subsamples of APC residues were collected from the different ash trucks and mixed into one 10-kg sample. Analysis of Heavy Metals. The Cd, Cr, Cu, Ni, Pb, and Zn contents in the ash and waste samples were analyzed by an AA320N atomic absorption spectrophotometer (AAS, Shanghai Analytical Instrument Overall Factory, China) after HCl/ HNO3/HF/HClO4 digestion. The concentration of heavy metals in the leachates was determined using AAS after concentrated HNO3 digestion. Calibration was performed every time an analysis was conducted. To prevent metal
FIGURE 2. Percentile contents of heavy metals in the major MSW compositions and incineration residues. Median, 10th, 25th, 75th, and 90th percentiles are plotted as the horizontal solid lines of the boxes. The dotted line represents the mean content value in each case. Outliers are shown as dots.
TABLE 1. Comparison of Heavy Metal Contents in MSW from the Literature with this Study waste
Cd
Cr
Cu
Ni
Pb
Zn
household waste (mg/kg) (9) 2(5 30 ( 23 289 ( 764 206 ( 360 160 ( 223 MSW (mg/kg) (28) 0.6-3.3 47-153 22-47 21-49 56-70 36-165 incinerated waste (mg/kg) (29–31) 2-22 20-105 335-1021 14-51 110-1500 759-1430 MSW compost (mg/kg) (1, 32–35) ND-7.0 35-209 10-600 6.5-149 3.4-800 110-1700 major MSW compositions 0.2-6.1 46-559 24-776 12-114 47-447 96-638 a in this study (mg/kg) (1.5-2.8) (143-229) (78-221) (30-51) (98-173) (228-323) a
Data range (the 95% confidence interval for mean heavy metal contents in MSW).
contamination, only guaranteed reagents (GR) were used. Each sample was analyzed in triplicate, accompanied by a digestion blank. Data Treatment and Statistical Analysis. SPSS 13.0 for Windows (27) was used to conduct statistical analysis, including descriptive statistics, independent-samples t test, one-sample t test, and one-sample Kolmogorov-Smirnov test. The Grubbs’ test was used to detect the outlying observations.
Results and Discussion Heavy Metal Contents in the Major MSW Compositions. The temporal variability in MSW composition and heavy metal contents in the MSW and incineration residues is shown in the Supporting Information. No obvious seasonal trends could be concluded, and the heavy metal contents in the incineration residues were more significantly variable than those in nonhazardous MSW, suggesting the influence of
hazardous components. Figure 2a-b presents the heavy metal contents (on a dry basis, the same below, unless indicated otherwise) in the major MSW compositions from the two incinerators. Large variability in the data was observed due to the great heterogeneous and temporal variable composition of the collected MSW. A similar heavy metal pattern was found for the two incinerators, where Zn, Cr, Cu, and Pb were the main heavy metals (average amount greater than 100 mg/kg) in the waste, followed by Ni (15-77 mg/kg) and Cd ( Pb ≈ Cu ≈ Cr > Ni > Cd; this distribution is similar to that in the MSW. However, volatile metals such as Cd, Pb, and Zn increased considerably in the APC residues (reaching to more than 10 times, 4 times, and 2 times larger than that in the bottom ash). In spite of the greater homogeneity of the incineration residues than MSW physically (24), a wide range of data was still observed, corresponding to the temporal change in the quantity and types of hazardous components 6214
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008
in MSW, physical composition of MSW, and operation conditions (such as incineration temperature, excess air rate, perturbation of air current and heat current, and lime and active carbon dosage for air pollution control). The data were within the reported values (3) for the heavy metals occurrence in the incineration residues. Mass Flows of Heavy Metals in the Incinerators. The results are illustrated in Figure 3a-f. The incinerated nonhazardous MSW compositions contained less than 4 g of Cd, 60 g of Ni, 250 g of Cr and Pb, 400 g of Zn, and 500 g of Cu for each ton of wet waste. Most of the dots representing the heavy metals Cd, Cu, Pb, and Zn were irregularly distributed above the diagonal, suggesting that these metals contained in the incineration products exceeded the contribution from nonhazardous MSW. The plots of Cr and Ni were more evenly displayed beside the diagonal, indicating the better closed level of the data. The heavy metal flows in
x /Fx FIGURE 4. FIP MSW for the two incinerators. Median, 10th, 25th, 75th, and 90th percentiles are plotted as the horizontal solid lines of the boxes. The dotted line represents the mean content value in each case. Outliers are shown as dots.
the nonhazardous MSW and incineration products were not significantly correlated. The ratio (R) of the heavy metal mass in the incineration products to that in the major MSW compositions (eq 6) reflects the closed level of the data. When R is significantly larger than 1, it indicates that the components other than the general MSW fractions (metals, glasses, vegetable and fruit waste, C&D debris, wood, textiles, plastics, and paper) significantly contribute to the occurrence of heavy metals in the waste. On the other hand, the closer the ratio is to 1, the better is the source separation of hazardous components applied. (R-1)/R indicates the contribution of the hazardous components to the heavy metals in the total MSW flows. R(-) )
x FIP x FMSW
,
x FH-MSW x FIP
)
x x FIP - FMSW x FIP
)
R-1 R
(6)
As shown in Figure 3g, even though the exhaust gas was not considered, R was basically in the range of 1-8. The high reproducibility of the mean ratio R pattern between the two incinerators (Figure 4) proved the reliability of the data. The mean values of R for Ni from the two incinerators were similar at a 0.01 significance level, and for the other heavy metals, the average ratios were statistically equal (significance >0.05), based on the independent-samples t test (Table 2.) Using the one-sample t test, the difference between the mean values of R was compared with 1, as summarized in Table 2. Only the average values of R for Cr and Ni were not statistically different from 1 (at a 0.05 significance level), suggesting that no significant Cr- and Ni-enriched pollutants were present in the major MSW compositions. The average values of R for the other metals were all significantly higher than 1. The results demonstrated that the sum of these heavy metals in the nine MSW fractions might underestimate the heavy metals flow in the MSW management system. Other minor components could contribute a significant part of the heavy metals in the collected MSW. Implication for the Hazardous Components Management. Some outlying observations were noticed in Figure 4. All the data except that for Cd (resulting from some extremely high values) were distributed normally according to the onesample Kolmogorov-Smirnov test (two-tailed significance >0.05). Therefore, the Grubbs’ test was conducted to detect and omit the outlying observations, using eq 7. The values larger than the critical value (37) for the Grubbs’ test were regarded as the outliers. The test can detect one outlier at a time Gn )
xmax - xj s
(7)
where Gn ) Grubbs’ value; xmax ) the suspected single outlier, i.e., the largest value; xj ) the mean value; and s ) the standard deviation. Three outliers for Cd (26.01, 22.67, 20.62), two outliers for Cu (11.66, 9.88) and Pb (8.37, 11.64), and one outlier for Ni
(7.62) and Zn (10.78) were finally omitted. Then, the 95% confidence intervals for the mean values of the remainder R were obtained by descriptive statistical analysis, i.e., 2.27-4.00 for Cd, 0.95-1.92 for Cr, 1.90-3.77 for Cu, 0.89-1.56 for Ni, 2.25-3.51 for Pb, and 2.98-4.06 for Zn. The contribution of household components to the heavy metals in the MSW could be estimated as 100% × (R-1)/R, which showed that 56-75% of Cd, 47-74% of Cu, 56-72% of Pb, and 66-75% of Zn in the waste could be possibly attributed to the minor fraction in MSWshazardous components. It was reported (20) that in Denmark in 1994, 25% of Pb in MSW originated from cathode ray tubes and another 25% from shredder fluff from accumulators, balance weights, solders, pigments, lead lining, etc. Sinkers for fishing, pigments in paint and plastic, and solders accounted for 11%, 8%, and 7% of Pb, respectively; in EU (2000), the Cd in discarded products disposed of with MSW was estimated to be 38%. Another study on the effect of spent batteries in EU countries to the heavy metals in MSW revealed the minor contribution of batteries, i.e., 2% of Zn from primary batteries, 2.7% of Ni and 17% of Cd from Ni-Cd batteries, and 0.4% of Pb from Pb-acid batteries. A higher contribution of batteries was estimated by Nakamura et al. (21) in Japan. In that study, small sealed lead batteries (27.4%), lead tubes (17%), crystal glassware (33.9%), printed circuit boards (6.5%), and electric bulbs (5.4%) contributed approximately 91% of Pb in MSW; 93.5% of Cd was contributed by Ni-Cd batteries (91.9%) and dry batteries (1.6%). Similar results for Cd were obtained by Heck et al. (23) in the U.S.A., that 91.3% of Cd in MSW was from batteries, but they estimated that batteries only contributed 0.43% of Pb. Electronic waste was also reported (5) to be the major hazardous component in MSW in Germany, accounting for about 46% of Pb, 55% of Cd, and 16% of Zn. Different consumption behavior and source collection practices for hazardous components in different countries result in the diverse distribution of hazardous components in the MSW. Furthermore, the different quantification methods (some were based on the production and discharging rates, some were based on spot check) may also yield varied results. In comparison with the contribution ratio of hazardous components from the literature, which varied from 38% to 93.5% for Cd and from 0.43% to 91% for Pb, moderate values were obtained in this study. In Shanghai city, separate collection of batteries has been practiced for several years; however, there are still some batteries mixed in MSW. Source separation of other hazardous components such as fluorescent lamps, printed circuit boards, and electronic scraps is not well established. The obtained contribution ratios provide for the first time the evidence of hazardous MSW components contamination in Shanghai, which is useful for applied waste management practices and future management strategy. The data indicate the need for further established and wider source separation of hazardous components in Shanghai and establish a VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6215
9
Mean difference is significant at a 0.05 level (2-tailed) when the significance
