Emissions of polychlorinated biphenyls (PCBs) from sludge drying beds to the atmosphere in Chicago

Available online at www.sciencedirect.com

Chemosphere 71 (2008) 1028–1034 www.elsevier.com/locate/chemosphere

Emissions of polychlorinated biphenyls (PCBs) from sludge drying beds to the atmosphere in Chicago Seung-Muk Yi a, Sandhya Reddy Pagilla b, Yong-Chan Seo c, William J. Mills d, Thomas M. Holsen b,* a

School of Public Health, Department of Environmental Health, Seoul National University, Yeongon-dong, Chongro-gu, Seoul 110-799, Republic of Korea b Department of Civil and Environmental Engineering, Clarkson University, Box 5710, Potsdam, NY 13699, United States c Department of Environmental Engineering, Sangji University, Wonju 220-702, Republic of Korea d Mills Consulting Inc., Oak Park, IL 60304, United States Received 8 August 2007; received in revised form 14 December 2007; accepted 30 December 2007 Available online 4 March 2008

Abstract Ambient air PCB concentrations in the Lake Calumet region in Southeast Chicago have been found to be significantly higher than in nearby non-urban areas. This area is highly industrialized and also contains municipal sludge drying facilities and landfills. In an effort to quantify the importance of the sludge drying facilities to the elevated concentrations, upwind/downwind air samples from the Calumet East sludge drying bed were obtained between April and October 2002. For these samples, the downwind minus upwind (downwind– upwind) concentration varied from 0.33 to 1.27 ng m3 for non-northeast (NE) direction winds suggesting sludge drying is a source of PCBs to the atmosphere. However, the upwind concentrations were higher than the downwind for winds from the NE of the sampling site suggesting more significant source(s), possibly Lake Calumet or the so called ‘‘Cluster site” NE of the sludge drying beds. Flux chamber experiments carried out during the sampling period measured average PCB fluxes of 210 ng m2 h1 (range 43–910 ng m2 h1) which resulted in an overall flux of 0.005 kg day (d)1 (2 kg yr1). A developed regression equation between moisture content and sludge concentration estimated higher PCB losses of 0.26 kg d1 (95 kg yr1). Although these two approaches yielded different values, they both indicate that the emission from the Calumet East sludge drying beds were of minimal importance when compared to the total estimated amount of 2–70 kg d1 (700–2100 kg yr1) of PCBs entering the Chicago atmosphere. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Air emissions; Air sampling; Fluxes; PCBs

1. Introduction Recent research has shown that PCB concentrations in the urban areas surrounding southern Lake Michigan are significantly higher than in nearby non-urban areas (approximately 2–3 ng m3 vs. <0.5 ng m3) (Keeler, 1994; Cotham and Bidleman, 1995; Hillery et al., 1997, 1998; Simcik et al., 1997). In addition, the PCB flux measured in Chicago is up to three orders of magnitude higher than in the non-urban areas (Holsen et al., 1991; Franz *

Corresponding author. Tel.: +1 315 268 3851; fax: +1 315 268 7985. E-mail address: [email protected] (T.M. Holsen).

0045-6535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.12.031

et al., 1998). Other work has shown that in Chicago there can be differences in PCB air concentrations even for relatively closely located (ca. 7 km) sampling sites (Basu et al., 2004). Additionally, PCB measurements made during the Lake Michigan Urban Air Toxics study (LMUATS) and the Atmospheric Exchange Over Lakes and Oceans Study (AEOLOS) were used with hybrid receptor models to determine source regions of PCBs in Chicago. Additional samples were obtained west of Chicago to help pinpoint sources. These results indicated areas (1) southwest between Joliet and Chicago, (2) northwest (NW) pointing towards Madison, WI and (3) the neighborhood of Lake Calumet, as PCB source areas. Six upwind/downwind

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

samples of the Calumet East sludge drying beds, obtained from 1998 to 2000, identified them as potential sources of PCBs (Hsu et al., 2003). These studies indicate that Chicago and other urban areas near the Great Lakes should be considered as major sources of PCBs to the Great Lakes. The objective of this study was to obtain better spatially oriented upwind/downwind samples, sludge samples and flux chamber samples from the Calumet East sludge drying facility to determine the magnitude of PCB air emissions from this facility and their importance to the Chicago air shed.

2. Experimental section 2.1. Sampling program 2.1.1. Short-term air sampling Hi-Vol air samplers (TE-PNY1123 Accuvol Modified PUF sampler, Tisch Environmental, Inc (Tisch, 2007)) were used, collecting operationally defined, atmospheric gas (using polyurethane Foam (PUF), Tisch Environmental, Inc.) and particulate phase (using glass fiber filter (GFF), Tisch Environmental) PCBs (Table 1). Four air samplers were deployed surrounding sludge drying cells at the Calumet drying beds (Fig. 1). The positions of these samplers were selected so that at least one of them could be classified as upwind and one as downwind most of the time depending on the wind direction. Samples were taken every 12 days between April and October 2002 (Table 2) on the same dates as the Integrated Atmospheric Deposition Network (IADN) (EC, 2002) sampling that occurred at another location approximately 19 km north (N). 2.1.2. Flux chamber measurements Flux chamber measurements were made on two different occasions during the summer of 2002 (Table 1). Six samples were obtained each time, three in the daytime and three in the nighttime to directly measure the PCB flux during drying. The flux chamber was designed according to the EPA User’s Guide (Kienbusch, 1986). The flux measurements were made by placing a small chamber over the drying sludge (Winegar and Keith, 1993) and drawing air out through a single PUF plug. Additional details can be found elsewhere (Pagilla-Reddy, 2004). The flux from the sludge was calculated using the mass of PCBs determined in the flux chamber air, the sampling time, and the area under chamber.

1029

2.1.3. Sludge samples Sludge grab samples (n = 20) were collected on five occasions before, during and after the drying season from various locations in the drying bed. 2.2. Sample extraction and analysis Analytical methods and techniques used for extraction and analysis of the PCB samples were based on US Environmental Protection Agency (USEPA) Methods (USEPA, 1997a, 1999, 2000a,b). Additional specific details are provided in the Quality Assurance Project Plan (QAPP)(Holsen, 2002) and other project publications (Pagilla-Reddy, 2004). Briefly, each GFF was cleaned in a muffle furnace and stored at 16 °C until sent to the field for sampling. The PUF plugs were cleaned by rinsing followed by soxhlet extraction and stored in glass (Mason) jars (precleaned in muffle furnace) and then stored at 16 °C until sent to the field for sampling. After sampling, the PUF plugs were returned to their mason jars, placed in coolers with blue ice and shipped to the laboratory at Clarkson University. All glassware used in the extraction of the samples was cleaned, covered with clean aluminum foil and stored in an environment free from contamination prior to use (USEPA, 2000a). PUFs and GFFs returned from sampling were extracted by Accelerated Solvent Extraction (ASE) (Dionex ASE 300) using hexane according to US EPA Method 3545 (USEPA, 1997b). Sludge was also extracted with ASE using a mixture of acetone/hexane (1:1) (Dionex, 1999). Sludge samples were weighed in pre-baked aluminum weighing pans and pre-cleaned anhydrous sodium sulfate (granular, 10–60 mesh, Fisher Scientific, Cat. #S415-1) was mixed with the sludge to remove any water present. To determine sludge water content, separate aliquots of the sludge samples were weighed in a tared aluminum weighing pan and then dried in an oven at 105 °C overnight (USEPA, 1996b). The samples were then allowed to cool in a desiccator to ambient temperature before weighing. Surrogate standards (surrogates) 3,5-dicholorbiphenyl (PCB 14), 2,3,5,6-tetrachlorobiphenyl (PCB 65), and 2,3,4,40 ,5,6-hexachlorobiphenyl (PCB 166) were spiked to each sample before the ASE extraction. Extracted samples were concentrated in a Zymark Turbovap LV evaporator to approximately 5 ml under ultra high purity (UHP) grade nitrogen gas. Eighteen molar sulfuric acid was added to the concentrated extract and vortexed for 1 min for the first clean up step (Cotham and Bidleman, 1995). The extracts

Table 1 Sampling overview Sample type

Sampling method

Sampling frequency

Sampling location

No. of samples

Ambient air Flux chamber Sludge

6 h PUF sampler Flux chamber Grab

Every 12 days 2 Intensives – 3 day and 3 night Before and after drying season

4 around the sludge bed Drying bed Drying bed

15  4 = 60 232=2 20

1030

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

Fig. 1. Calumet Sludge Drying beds and air sampler locations.

were then fractioned on an alumina–silica gel column (1 cm sodium sulfate, 3 g 6% deactivated silica gel, 1 cm sodium sulfate, 10 g 10% deactivated alumina, and 1 cm of sodium sulfate). Columns were eluted three times with 25 ml hexane. The cleaned extracts were again concentrated to approximately 5 ml in the Zymark Turbovap LV evaporator. Approximately 2 g of activated copper powder (40 mesh) was added to the cleaned extracts to remove sulfur (USEPA, 1996a). Internal Standards (IS) 2,4,6-trichlorobiphenyl (PCB 30), 2,20 ,3,4,40 ,5,6,60 -octachlorobiphenyl (PCB 204), and tetrachloro-m-xylene (TCMX) were added to the sample before further concentrating to exactly 1 ml under gentle UHP nitrogen gas. The sample was then transferred to an autosampler vial. Sample extracts were analyzed on a Hewlett Packard (HP) 5890A GC with a 63Ni electron captured detector (ECD), a HP 7673A automatic sampler and a DB-XLB Capillary column (60 m, 0.25 mm i.d., 0.25 lm film thickness). Samples were injected in splitless mode and analyzed using a GC oven temperature program (initial temperature = 100 °C hold for 1 min, 1.75 °C min1 to 280 °C hold for 15 min, total run time  120 min). Injector and detector temperatures were 300 and 330 °C, respectively. All flux chamber sludge samples and at least 10% of the air samples

were confirmed with a DB-5 capillary column (60 m, 0.25 mm i.d., and 0.25 lm film thickness) (USEPA, 1997a) with the same temperature program as for the DB-XLB. Calibration standards (Accustandard Inc. mix of five individual solutions, 144 congeners) at 10 lg ml1), surrogates (PCB 14, 65 and 166), and IS (TCMX, PCB 30 and 204) were used (a total of 150 congeners including the surrogate and IS, some co-eluting). A three-point calibration curve, to monitor for any change in the retention time and detector response, was generated for each batch of samples analyzed with a single-point calibration daily as set out in the QAPP (Holsen, 2002). The r2 for the linear regressions of the calibration curves were greater than 0.999 for most of the PCB congeners. A number of QA/QC procedures were implemented for this work including, initial method validation, analysis of SRMs, matrix spikes, interlaboratory comparisons, method and field blanks, recoveries of surrogate standards, and breakthrough studies. During the initial air sample analyses, a couple of peaks (retention time (RT): 78 and 92 min on DB-XLB) were found in all samples including the lab blank. Peak areas were very small in the lab blanks but much higher in the

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

1031

Table 2 Average temperatures and wind speed for sampling dates Category

Date (mm/dd/yy)

Average Temperature (°C)

Wind speed (m/s)

Sampler 1 (ng/ m3)

Sampler 2 (ng/ m3)

Sampler 3 (ng/ m3)

Sampler 4 (ng/ m3)

Non-NE winds

5/9/02 5/26/02 6/19/02 7/1/02 7/26/02 9/23/02

14 18 26 30 27 18

8.38 3.45 4.34 2.94 2.57 3.38

0.11 1.05 2.19 3.52 1.10 1.12

1.27 0.70 2.35 2.87 1.06 1.60

0.62 0.74 1.71 1.93 0.83 0.73

0.49 0.70 0.84 2.25 1.02 0.18

NE winds

5/14/02 7/13/02 9/11/02

10 24 20

2.35 4.12 4.48

1.95 (U) 5.51 (U) 2.28 (U)

0.80 (U) 1.61 (U) 2.21 (U)

0.84 (D) 1.88 (D) 1.76 (D)

1.44 (D) 1.62 (D) 1.21 (D)

Mixed

8/30/02 10/7/02

25 10

2.57 1.91

0.63 (U) 2.29 (U)

2.18 (U) 1.01 (U)

1.98 (D) 0.55 (U)

1.26 (D) 1.15 (D)

No impact

6/7/02 8/18/02 11/7/02

19 21 7

2.94 1.69 3.67

1.44 (D) 2.07 (D) 0.55 (D)

2.15 (D) 2.40 (D) 0.38 (D)

1.58 (U) NV NV

1.23 (U) NV NV

(D) (D) (D) (D) (D) (D)

(D) (NA) (D) (D) (D) (D)

(U) (U) (U) (U) (U) (U)

(U) (U) (U) (NA) (NA) (U)

U = upwind, D = downwind, NV = data not valid, NA = neither upwind/downwind.

method blank and PUF samples. Polybrominated diphenyl ethers (PBDE) contamination was confirmed by GC-MS analysis (William Mills, personal communication) possibly due to the PBDE addition to the PUF during the manufacturing process (Jim Pagano, personal communication). As a result of the QA/QC studies above, some congeners which might be affected by PBDE (Alaee et al., 2001), and/ or some congeners showing signs of contamination were not included in the analysis (32 out of 144;PCBs 1, 2, 3, 12, 15, 16, 17, 18, 29,31,40, 56, 69, 77, 81, 84, 87, 103, 105, 117, 131, 134, 135, 136, 138, 144, 147, 167, 170, 189, 199, and 208). In our experience these congeners were not typically analyzed in other studies, or were not major components in Chicago air samples. Interlaboratory studies indicated good agreement for congeners that were common between laboratories. Surrogate recoveries for the samples ranged between 60% and 125%. The instrument detection limit (IDL) (determined as three times the standard deviation (SD) of a series of injections with a low concentration standard.) for the individual congeners ranged from 0.003 to 6.1 ng (Hsu, 2001). Breakthrough (from PUF1 to PUF2 (backup PUF)) was not found to be a significant issue. For a dual column analysis, if the values were within 20%, the average value was calculated and reported as confirmed. However, if the differences were >20% the lower number of the two values produced was used as the confirmed value. Any congener, which had an unusually high value when compared to its counterpart, was marked as unconfirmed. Further information on the QA/QC studies is provided elsewhere (Pagilla-Reddy, 2004). 3. Results and discussion 3.1. Ambient air samples In this study, sets of samples were obtained on 14 different dates in 2002 from around a sludge drying cell (Fig. 1)

of the Calumet plant of MWRDGC. Based on the wind direction during sampling, samples were divided into four categories; upwind/downwind, NE wind, mixed, and no impact (Table 2). The overall average PCB values for these air samples were 1.47 ng m3 (range 0.11–5.51 ng m3). These values were similar to the average PCB air concentration of 2.45 ng m3 (range 0.76–5.47 ng m3) from a previous, more limited study (Hsu et al., 2003). 3.1.1. Non-NE wind direction There are six sets of samples that were classified as upwind and downwind samples when the wind was not from the NE direction (Table 2). The average PCB downwind and upwind concentrations for these samples varied from 1.05 to 3.52 ng m3 and from 0.46 to 1.43 ng m3, respectively (see Supplementary Materials). The sample dated 05/09/02 is not included in the average concentration as the sludge beds were empty during this date. However even for this sample it appears that there is still some volatilization of PCBs from the drying bed surface. Both upwind and downwind concentrations increased with increasing temperatures. The average downwind minus upwind concentration was 1.11 ng m3 (n = 5) compared to an average downwind–upwind from previous study of 1.91 ng m3 (n = 3) (Hsu, 2001). The average downwind– upwind profile (see Supplementary Materials). indicated that there was an increase in the concentrations of both low and high molecular weight congeners. 3.1.2. NE wind direction The three sample dates classified as NE winds had prevailing wind coming from the NE direction (Table 2). For this group, the upwind concentrations were higher than the downwind concentrations (Fig. 2) suggesting that the upwind samples were affected by an unknown source(s) in the NE direction. The average downwind concentrations for these samples ranged from 1.14 to 1.75 ng m3 whereas the average upwind concentrations ranged from 1.38 to

PCB Concentrations (ng m-1)

1032

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

4

upwind downwind

3.5 3 2.5 2 1.5 1

congener pattern is similar to the congener distribution observed in the sludge) and corrected for the large organic carbon content (OC) in the sludge, the measured and modeled results agreed remarkably well (Fig. 4). The flux chamber congener profile was variable but in general enriched in lower chlorinated congeners relative to the sewage sludge (see Supplementary Materials).

0.5 0

05/14/02

07/13/02

09/11/02

Date

Fig. 2. NE wind direction – upwind vs. downwind PCB concentrations.

3.56 ng m3. Northeast of the site is Lake Calumet and the Southeast Chicago ‘‘Cluster Site” (IEPA, 1986; USEPA, 2006a,b), both of which are known to contain PCBs. The average downwind–upwind congener profile indicates that for most congeners the average upwind concentrations were higher than the average downwind concentrations (see Supplementary Materials). This contrasts with the samples that appear to be influenced by sludge drying (Pagilla-Reddy, 2004). 3.2. Flux chamber measurements The average PCB flux measured (n = 12) with the flux chamber was 210 ng m2 h1 (range 43–910 ng m2 h1). This value is slightly smaller than the values estimated with K-theory (250–970 m2 h1), using previously measured upwind and downwind concentrations (Hsu et al., 2003). Theoretical fluxes were calculated using an equation developed for soils, (y = 2130.7x0.6542) (Mills et al., 2004) where x is sludge concentration (mg kg1 dry) and y is flux (ng m2 h1). Modeled fluxes were 2–12 times higher than the measured fluxes with the greatest difference at the lowest temperature and smallest difference when the temperature was close to 25 °C; the temperature the equation was developed for. When the measured fluxes were corrected to 25 °C using the relationship between temperature and Aroclor 1260 Pvap developed by Mills et al. (Mills et al., 2004) (this assumption is reasonable since Aroclor 1260

3.3. Sludge samples Twenty grab samples were collected during this study (April to October 2002) from all stages of the sludge drying process. The average concentration was 2.0 mg kg1 dry (range 0.79–3.1), a value somewhat higher than previous results for this site (0.6–1.0 mg kg1 dry) (Hsu et al., 2003). In general the sludge contained higher concentrations of high molecular weight congeners than low molecular weight congeners, although the concentrations were highly variable (see Supplementary Materials). High molecular weight congeners have lower Pvap and larger partition coefficients, so they should be more strongly associated with the sludge (due to its high OC content) than the low molecular weight congeners. The 20 sludge samples were segregated in three different moisture contents ranges from minimum to maximum (>70%, 70–20%, and <20%) to determine how the concentration and congener profile change during drying (see Supplementary Materials). The correlation observed for moisture vs. PCB concentration was quite good (r2 = 0.93) (Fig. 3). Although the slope of this regression line found in this study (0.0125) is similar to that found in a previous lab experiment (0.0081) (Hsu, 2001), the PCB concentrations measured as part of this study were significantly higher than in the lab experiment. 3.4. Comparison of PCB congener profiles in the sludge, air, and flux chamber experiments The average sludge congener profile was fairly evenly distributed between low and high molecular weight 4000

measured flux, actual conditions 3500

measured flux(T and OC adjusted)

October 8-9

PCB flux (ng m-2 hr-1)

predicted flux (T and OC adjusted) 3000

2500

August 20-21

2000

1500

1000

500

0

0.94

1.8

PCB concentration (mg/kg, dry)

Fig. 3. Relationship between sludge concentration and moisture content in Calumet Sludge Drying beds.

Fig. 4. Measured vs. predicted average fluxes for two different dates. 10% error bars shown for reference purposes.

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

congeners, however there were fewer very low molecular weight congeners present than were found in the ambient air or in the flux chamber effluent samples (see Supplementary Materials). However the apparent relative enrichment in mid and high molecular weight congeners in the non NE downwind–upwind air samples relative to the flux chamber samples is probably not significant due to the large error bars associated with these measurements (see Supplementary Materials) and the different scale of these measurements (small flux chamber area vs. the large area influence of the downwind and upwind samples). 3.5. Sludge PCB emission inventory The total emission rate of PCBs from the Calumet sludge drying beds was estimated to be 0.003 kg d1 (1 kg yr1) using the average flux measured in the flux chamber experiment (210 ng m2 h1, (range 43– 910 ng m2 h1) times the total estimated area of 514,000 m2). If both Calumet and Stickney sludge beds sludge drying bed areas are used (1,100,000 m2) to estimate the total flux, a flux of 0.005 kg day1 (2 kg yr1, range 0.4–9 kg yr1) is estimated. This value is in fair agreement with the K-theory values for these sources made by Hsu et al (Hsu et al., 2003) (0.0074 and 0.0078 kg d1), but is an order of magnitude lower than their results from an EPA virtual point source model (0.059 and 0.053 kg d1). Another approach can be used to calculate the emission rate from the sludge beds using the relationship derived between the moisture content and the sludge concentrations (Fig. 3). Y ¼ 0:0125X þ 1:4056ðr2 ¼ 0:93Þ

ð1Þ 1

where Y = PCB concentrations (mg kg dry) and X = moisture content. The MWRDGC Calumet East sludge drying beds initially contain sludge at 75–85% moisture content that is reduced to approximately 35% before leaving the sludge facility. Substituting a moisture content of 35% in Eq. (1), the dried sludge PCB concentration can be estimated to be 1.8 mg kg1 dry. Assuming the initial moisture content of 80% yields an initial PCB concentration of 2.4 mg kg1 dry. The percentage of PCB volatilized would be 25%. Assuming the annual sludge production was equal to that in 2000, when the Calumet East drying beds dried 170,685 dry tons (in total 232,489 dry tons from all seven MWDRGC plants) (Tata, 2000), PCB volatilization from Calumet East drying beds and from all seven MWRDGC plants would be 0.26 kg d1 (95 kg yr1) and 0.36 kg d1 (131.4 kg yr1), respectively. These values are very similar to earlier estimates made with lab experiments (Hsu et al., 2003). Therefore both the flux chamber approach and sludge sampling approach yield PCB emission estimates that show this source represents only a small fraction of the recent estimates of PCB emissions in Chicago of 2–70 kg d1 (700–25,000 kg yr1) (Hsu et al., 2003).

1033

4. Conclusions The downwind–upwind concentrations of PCBs near the sludge drying beds varied from 0.33 to 1.3 ng m3 when the winds were not blowing from the NE direction. The average value of 1.1 ng m3 is somewhat smaller than that found in a previous study (1.9 ng m3) in 1999 and 2000 (Hsu et al., 2003). The downwind–upwind congener profile (non-NE) contained fairly evenly distributed high and low molecular weight congeners indicating PCB losses from the sludge in all molecular weights. However, the upwind concentrations were higher than the downwind concentrations in cases where the wind direction was from the NE of the sampling site. This may be as a result of other PCB sources present to the NE of the sampling site, possibly Lake Calumet itself and/or the Cluster site and/or other sites. Flux chamber measurements were conducted twice during the sampling period. The average measured flux was 210 ng m2 h1 (range 43–910 ng m2 h1). The measured flux is slightly lower than the flux estimated from K-theory and agreed fairly well with the flux predicted with a published soil flux model (Mills et al., 2004) when temperature and OC content of the sludge was taken into account. The average concentration of PCBs in the sludge samples were 2.0 mg kg1 dry (0.79–3.1 mg kg1 dry) somewhat higher than values in a previous study. The sludge samples were more enriched with high molecular weight congeners than in low molecular weight congeners than typically found in the air. The high correlation (r2 = 0.93) between the sludge concentration and moisture content suggests a significant loss of PCBs during drying, probably due to volatilization. Results of this study were used to estimate the emission rate from all MWRDGC sludge beds of 2 kg yr1 (0.006 kg d1) (Flux Chamber) and 132 kg yr1 1 (0.35 kg d ) using a regression equation. Both these estimates indicate that the emission from the Calumet East sludge drying beds were not significant compared to the total PCB that was estimated to enter the Chicago atmosphere using a simple box model 700–25,000 kg yr1 (2–70 kg d1).

Acknowledgements This work was funded in part by the USEPA Great Lakes National Program Office, Todd Nettesheim Project Officer. We thank Will Stefan, Eleanor Hopke and YingKuang Hsu for all their assistance. The cooperation of the Metropolitan Water Reclamation District of Greater Chicago is greatly appreciated. Mention of any company or trade names is for informational purposes and does not constitute endorsement by any of the organizations or the authors. This is the Clarkson Center for the Environment publication No. 346.

1034

S.-M. Yi et al. / Chemosphere 71 (2008) 1028–1034

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.chemosphere. 2007.12.031. References Alaee, M., Backus, S., Cannon, C., 2001. Potential interference of PBDEs in the determination of PCBs and other organochlorine contaminants using electron capture. J. Sep. Sci. 24, 465–469. Basu, I., Hafner, W.D., Mills, W.J., Hites, R.A., 2004. Differences in atmospheric persistent organic pollutant concentrations at two locations in Chicago. J. Great Lakes Res. 30, 310–315. Cotham, W.E., Bidleman, T.F., 1995. Polycyclic aromatic hydrocarbons and polychlorinated biphenyls in air at an urban and rural site near Lake Michigan. Environ. Sci. Technol. 29, 2782–2789. Dionex, 1999. Accelerated solvent extractor (ASE) of polychlorinated biphenyls from polyurethane foam adsorbent cartridges. Appl. Note, 333. EC, 2002. Integrated Atmospheric Deposition Network Environment Canada, Toronto. Available at . Franz, T.F., Holsen, T.M., Eisenreich, S.J., 1998. Dry deposition of particulate polychlorinated biphenyl and polycyclic aromatic hydrocarbons to Lake Michigan. Environ. Sci. Technol. 32, 3681–3688. Hillery, B.R., Basu, I., Sweet, C.W., Hites, R.A., 1997. Temporal and spatial trends in a long-term study of gas-phase PCB concentrations near the Great Lakes. Environ. Sci. Technol. 31, 1811–1816. Hillery, B.M., Simcik, M.F., Basu, I., Hoff, R.M., Stachan, W.M.J., Burniston, D., Chan, C.H., Brice, K.A., Sweet, C.W., Hites, R.A., 1998. Atmospheric deposition of toxic pollutants to the Great Lakes as measured by the integrated atmospheric deposition network. Environ. Sci. Technol. 32, 2216–2221. Holsen, T., 2002. Quality Assurance Project Plan (QAPP), Project Title: sources of Polychlorinated Biphenyls (PCBs) to the atmosphere in Chicago. Clarkson University, Potsdam, NY, p. 29. Holsen, T.M., Noll, K.E., Liu, S.P., Lee, W.J., 1991. Dry deposition of polychlorinated biphenyls in urban areas. Environ. Sci. Technol. 25, 1075–1081. Hsu, Y., 2001. PhD Dissertation. Civil and Environmental Engineering, Clarkson University, Potsdam, NY. Hsu, Y., Holsen, T., Hopke, P., 2003. Locating and quantifying PCB sources in Chicago: receptor modeling and field sampling. Environ. Sci. Technol. 37, 681–690. IEPA, 1986. The Southeast Chicago study: an assessment of Environmental Pollution and Public Health Impacts. Environmen-

tal Programs, Illinois Environmental Protection Agency, Springfield, IL. Keeler, G.J., 1994. The Lake Michigan Urban Air Toxics Study, Final report. Atmospheric Research and Exposure Assessment Laboratory, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC. Kienbusch, M., 1986. Measurement of gaseous emissions rates from land surfaces using an emission isolation flux chamber, EPA Users Guide. Mills, W., Bennett, E., Schmidt, C., Thibodeaux, L., 2004. Obtaining quantitative vapor emissions estimates of polychlorinated biphenyls and other semivolatile organic compounds from contaminated sites. Environ. Toxicol. Chem. 23, 2457–2464. Pagilla-Reddy, S., 2004. Loss of Polychlorinated Biphenyls (PCBs) from sewage sludge during the drying process. Department of Civil and Environmental Engineering. Clarkson, Potsdam, NY. Simcik, M., Zhang, H., Eisenrich, S., Franz, T., 1997. Urban contamination of the Chicago/Coastal Lake Michigan atmosphere by PCBs and PAHs during AEOLOS. Environ. Sci. Technol. 31, 2141– 2147. Tata, P., 2000. 1995 to 1999 organic compound concentration in effluent and biosolids report from metropolitan water reclamation District of Greater Chicago. Response to FOIA Request #00-207 by Ying-Kuang Hsu. Metropolitan Water Reclamation District of Greater Chicago. Tisch, 2007. TE-PNY1123 Accuvol Modified Puf Sampler. Tishch Environmental. USEPA, 1996a. Method 3660B, Sulfur cleanup. SW-846 On-line. OSWER, Washington, DC, p. 6. USEPA, 1996b. SW-846 test methods for evaluating solids waste. Physical/chemical methods, Method 8275A. Offices of Solid Waste and Emergency Response, US Environmental Protection Agency, Crystal City, VA. USEPA, 1997a. Lake Michigan Mass Balance Study (LLMB) Methods Compendium. USEPA, 1997b. SW-846, Update III: Test methods for Evaluating Solid Wastes, Method 3545:. U.S.GPO. USEPA, 1999. Compendium of Methods for Determination of Toxic Organic Compounds in Ambient Air, second ed. Compendium Method TO-4A. USEPA, 2000a. Method 8082a, Polychlorinated Biphenyl’s (PCBs) by Gas Chromatography. USEPA, 2000b. Test methods for evaluating solid waste. Physical/ chemical methods, SW846 Online. Office of Solid Waste and Emergency Response, Environmental Protection Agency, Washington, DC. USEPA, 2006a. Lake Calumet cluster site. USEPA, 2006b. NPL Site Narrative for Lake Calumet Cluster. Winegar, E.D., Keith, L.H., 1993. Sampling and Analysis of Airborne Pollutants. CRC Press, Boca Raton, Florida.