Spatial Distribution of Atmospheric PCBs in Zurich, Switzerland: Do

Article pubs.acs.org/est

Spatial Distribution of Atmospheric PCBs in Zurich, Switzerland: Do Joint Sealants Still Matter? Pascal S. Diefenbacher,†,‡ Andreas C. Gerecke,*,‡ Christian Bogdal,†,§ and Konrad Hungerbühler† †

Institute for Chemical and Bioengineering, ETH Zurich, Vladimir-Prelog-Weg 1, CH-8093 Zürich, Switzerland Empa, Swiss Federal Laboratories for Materials Science and Technology, Ü berlandstrasse 129, CH-8600 Dübendorf, Switzerland § Agroscope, Institute for Sustainability Sciences ISS, Reckenholzstrasse 191, CH-8046 Zürich, Switzerland ‡

S Supporting Information *

ABSTRACT: Passive air samplers were deployed at 23 sites across the city of Zurich, Switzerland to investigate the spatial distribution of polychlorinated biphenyls (PCBs) in air. Concentrations of the six indicator PCBs (iPCBs) in air ranged from 54 to 3160 pg·m−3 in the two sampling campaigns (spring 2011 and spring 2013). Measurements at two sampling sites were significantly higher than the median in both years, because of the proximity to primary PCB sources. Concentrations at most other stations were in a narrow range, suggesting that atmospheric PCB concentrations in Zurich are mainly caused by a high number of rather small sources. A correlation of iPCB concentrations in air with the number of buildings constructed between 1955 and 1975 in the surrounding areas of the sampling sites was observed. This demonstrates that PCB-containing building materials, such as joint sealants, influence PCB levels in urban air. Additionally, atmospheric iPCB concentrations were measured in the surrounding of a housing complex with PCB-contaminated joint sealants. Using a Gaussian diffusion model, annual iPCB emissions of 110−190 g were calculated for this housing complex. This appreciable amount released by a single building points out that more efforts are required to further eliminate remaining PCB stocks.

■

INTRODUCTION Polychlorinated biphenyls (PCBs) are a well-known class of persistent organic pollutants (POPs) that were extensively used from the 1950s to the 1970s. Due to their high chemical stability and low flammability, PCBs were used in a wide range of applications, mainly as dielectric fluids in transformers and capacitors, and as plasticizers in rubber, plastics, sealants, paints, and adhesives.1 PCBs are persistent in the environment, toxic, and bioaccumulative and, therefore, pose a risk to human health and biota. Switzerland banned the use of PCB in open applications such as sealants, paints, adhesives, and lubricants in 1972. A complete ban of PCBs in all applications including closed systems (e.g., transformers and capacitors) took effect in 1986 and worldwide through the Stockholm Convention on POPs in 2004. Whereas PCBs in closed applications can be inventoried, recovered, and disposed of more efficiently, PCBs in open applications containing application are considerably more difficult to track and, therefore, continue to be in use for decades. Thus, PCBs can volatilize from large reservoirs in PCB-containing materials and open equipment still in use. Elevated concentrations of PCBs in air in highly populated areas and urban plumes contributing to the environmental contamination by PCBs in surrounding regions has been shown in several studies.2−6 Also the distribution of PCBs in the urban environment has been examined.7−12 Due to the variety of possible atmospheric sources within urban areas, they are still not well characterized. Some studies supposed that PCB emissions in cities are caused by a number of small diffuse sources.9,10 However, substantial gaps remain in the under© XXXX American Chemical Society

standing of the relative importance of sources type and the distribution of atmospheric PCB levels in urban areas. This study focuses on the city of Zurich, Switzerland, representing a typical urban region in an industrialized country. Additionally, we can profit from our previous studies, where we could show that Zurich still represents an important source region for PCBs with iPCB emissions of 13 kg per year for the entire city.13−15 Furthermore, we wanted to overcome the main limitation of our previous studies, which were all based on PCB measurements at single locations and the resulting emissions were derived under the assumption that PCBs are homogeneously distributed in environmental compartments. In this study, PCB concentrations in air were determined in spring 2011 and spring 2013 at 23 sites distributed in the city of Zurich to assess the intraurban variability. We further improve the characterization of sources by answering the question if PCBs are emitted from a large number of small sources or from a few large sources. It is known that a significant use of PCBs was in joint sealants of building constructed from the 1950s to the 1980s. Kohler et al.16 declared joint sealants as an “overlooked diffuse source of PCBs”. Also Robson et al.17 indicated that joint sealants represent a reservoir of PCBs that can last for decades. To reconsider these statements, we calculated the source strength Received: September 21, 2015 Revised: November 11, 2015 Accepted: November 18, 2015

A

DOI: 10.1021/acs.est.5b04626 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 1. Atmospheric iPCB concentrations measured in spring 2011 (a) and 2013 (b) in the city of Zurich, Switzerland. (c) and (d) show the boxand-whisker plot stretching from the 25th-percentile to the 75th-percentile, with the median line in between. Lower whisker represents lowest values within −1.5 interquartile range of the first quartile, upper whisker cover +1.5 interquartile range of the third quartile. Outlying air concentrations are represented by black diamonds.

PCBs was eluted with 15 mL DCM/Hex (1/1) from the first column and with 15 mL Hex from the second column. The final extract was solvent exchanged to toluene and reduced to approximately 20 μL. Further, 2 ng of 13C12-labeled-PCB-70 was added as recovery standard prior to analysis. PCB concentrations were determined by gas chromatography coupled to electron ionization high resolution mass spectrometry (GC/EI-HRMS) on a Thermo Finnigan MAT 95 high resolution mass spectrometer. A detailed description of the analytical method was presented in Diefenbacher et al.15 Study Design. To determine the intraurban distribution of PCB concentrations in air, PAS were deployed at 23 locations across the city of Zurich. The first campaign with a deployment time of 29 to 42 days was performed in spring 2011. A second campaign with identical sampling sites and a deployment time of 40 to 49 days was performed in spring 2013. In parallel to the analysis of PCB, short-chain chlorinated paraffins were determined in these samples as reported previously.21 In both years, seven sites (A1-A7) were located in the less-dense populated suburban areas to gain information on the PCB background concentration in urban air and further seven sites (B1−B7) were distributed within the urban area to account for intraurban deviations. The remaining nine PAS (C1−C9) were situated in a specific neighborhood, to observe the local distribution of PCB concentrations in air with a higher spatial resolution. This specific neighborhood was selected because it exhibited a lot of construction activities in the 1950s to 1980s, which is the period of time when PCBs were mainly used. Additionally, iPCB concentrations were determined at one rural sampling station (Tänikon, 30 km away from the city center) in spring 2013. A detailed description of all sites and a map can be found in the Supporting Information. Air temperature, wind speed, and rainfall rates were calculated by averaging measurements from three nearby meteorological stations. Besides the examination of the intraurban distribution of iPCBs this works aims to assess the importance of PCBcontaining building materials for the total urban emissions. Therefore, further air measurements were performed next to a

of a single housing complex with PCB-contaminated joint sealants.

■

MATERIALS AND METHODS

PCB Passive Air Sampling. Passive air samplers (PAS) consisting of polyurethane foam (PUF) disks housed in stainless steel chambers were employed to study concentrations of iPCBs in urban air. A detailed description of this sampling method was presented in Diefenbacher et al.15 All PUF disks were cleaned by Soxhlet extraction with dichloromethane/nhexane (DCM/Hex) (1/1) for 12 h. Subsequently, they were wrapped in clean aluminum foil and stored at −20 °C in sealed polyethylene plastic bags until deployment. PAS were installed at various locations within the city of Zurich, either on the ground of open spaces or on roof tops of buildings. A requirement for all sampling sites was that ambient air could circulate freely around the sampler; hence, they were installed on a height of 1.5−3 m above the specific ground and at least 3 m from a building wall. After the deployment, samples were wrapped in clean aluminum foil and stored at −20 °C until analysis. To derive air concentrations, a uniform sampling rate (Rs) of 4 m3·d−1 was used for all PCB congeners, which is consistent with previous studies.18−20 Additionally, a calibration experiment with a high-volume air sampler situated next to a PAS was carried out to prove that this Rs value is applicable for our sampling setup.15 Sample Extraction and Analysis. Target analytes in this study were the iPCBs (PCB-28, -52, -101, -138, -153, and -180). Prior to extraction, 5 ng of 13C12-labeled iPCB analogues were spiked to the PUF as internal standards. The PUFs were Soxhlet extracted for 12 h using DCM/Hex (1/1). The extracts were cleaned by chromatography through a column packed with anhydrous Na2SO4 (2 g), activated silica gel (2.8 g), and acidic silica gel (1.7 g, 44% concentrated sulfuric acid). A second column containing deactivated Florisil (3 g, 1.5% water) topped with anhydrous Na2SO4 (2 g) was used to separate PCBs from other organic pollutants. The fraction containing B


Article


Even though the median PCB concentrations in air were quite similar during both sampling campaigns, it is obvious that air concentrations measured in the second campaign showed much more variance. In 2011, the maximum value measured at site C8 was higher than the median by approximately a factor of 4, whereas in 2013 this difference accounted for approximately a factor of 20. This may be explained by different wind conditions. Although average wind speeds were around 1.5 m· s−1 for both years and the prevailing wind directions were similar, these wind data were derived from only three meteorological stations within the city of Zurich and, thus, cannot account for differences in the complex wind regime on the scale of a single PAS sampling site, such as C8. The congener pattern observed at different sites in Zurich exhibited a high variability and was dominated by lower chlorinated congeners at most sites as expected for volatilization sources (Figure S2 and S3). However, a sound interpretation of PCB congener patterns and the attribution to specific emission sources would require the analysis of more PCB congeners. A comparison of PCB concentrations in air with other urban areas is limited due to differences in the set of target compounds. However, a study of the distribution of PCBs in the atmosphere of Toronto, Canada observed levels ranging from 6 to 1300 pg·m−3 (sum of 86 PCB congeners) at 19 sites within the city.12 Even more variation was found in Chicago, USA, where concentrations of 75−5500 pg·m−3 (sum of 169 PCBs) were measured.10 Further, PCB masses in PAS showed strong spatial gradients and varied over a factor of 30 in the metropolitan area of Philadelphia, USA.7 Due to the high variability of iPCB levels in urban areas, a careful selection of sampling sites is required to achieve representative concentrations in air. Previous studies that aimed to determine total PCBs emissions in Zurich were based on measurements at one single location (corresponding to the site B1 in this study).14,15 In this study, iPCB concentrations measured at site B1 were 80 pg·m−3 and 110 pg·m−3 in 2011 and 2013, respectively. Hence, they were close to the baseline concentration in the city of Zurich and it is probable that they are not directly influenced by nearby point sources. Thus, estimating emissions for the city of Zurich based on measurements at this location was a reasonable approach followed in our previous studies. Spatial Distribution of iPCB Concentration. A measurement of iPCB concentrations in spring 2013 at Tänikon, a rural sampling site with 30 km distance to the city center of Zurich exhibited 37 pg·m−3. Hence, these rural concentrations were 4 times lower than the urban median concentration measured during the identical time period and almost 3 times lower than the urban baseline concentration that was arbitrarily defined as the 25th percentile of all measurements in the city of Zurich. Such an urban-rural gradient was already examined in several studies that reported ratios of urban to rural concentrations of up to a factor of 40.3,4,12,22,25 However, these ratios cannot be generalized, due to the heterogeneity of PCB levels and the limited number of sampling sites within the study areas. Furthermore, the definition of a rural or background region is unclear and was not consistent in these studies. The observation that iPCB levels in the urban region are usually higher than levels in surrounding rural areas supports our assumption of significant PCBs emissions in the city of Zurich, which we suppose to originate from many small and localized sources.

housing complex consisting of two separate buildings with a total of 130 apartments on an area of 18 000 m2. These concrete buildings were constructed from 1967 to 1969 and were equipped with PCB-containing joint sealants between the concrete blocks. To determine the source strength of this individual point source, the trend of iPCB concentrations in air with increasing distance from the source was investigated. In spring 2015, 10 PAS were installed for a sampling period of 36 days in the surrounding environment. The sampling type and analysis was equal to the method described above. In this sampling campaign 9 PAS were attached to street lamps at a height of 3 m, whereas one PAS was installed on the roof top of an elementary school building, which corresponds to the site C4 in the intraurban campaigns.

■

RESULTS AND DISCUSSION

Atmospheric iPCB Concentrations in 2011 and 2013. In the year 2011, atmospheric iPCB concentrations measured at 23 sampling sites within Zurich showed an interesting distribution (Figure 1). Most of the sites (n = 19) exhibited iPCB concentrations in a narrow range below 250 pg·m−3, whereas concentrations determined at four sampling sites were significantly higher (544−723 pg·m−3; overall min-max 80 to 723 pg·m−3, median: 172 pg·m−3). During the second campaign in 2013 a much wider range of concentrations from 54 to 3160 pg·m−3 (median: 154 pg·m−3) was observed. iPCB concentrations at most sites (n = 14) were below 250 pg·m−3, seven stations exhibited concentrations between 250 and 850 pg m−3 and iPCB concentrations at two stations exceeded 3000 pg·m−3. Overall, the concentrations were in the range usually observed in cities in Europe and North America.10,12,22,23 The results show that besides single sites with remarkably high PCB concentrations, most sampling sites exhibited concentrations that were in a quite narrow range. We assume that these sites are not directly influenced by nearby PCBs sources and represent the baseline for atmospheric PCB concentration in the city of Zurich. As these levels are all close to the 25th percentile, this value was defined as the urban baseline concentration (138 pg·m−3 and 100 pg·m−3 in 2011 and 2013, respectively). Environmental conditions during these two sampling campaigns were comparable. Average air temperatures were 11.5 and 9.6 °C and wind speeds accounted for 1.4 and 1.5 m· s−1 in spring 2011 and 2013, respectively. The difference in urban baseline concentrations between the two campaigns is probably due to the differences in average air temperatures during the sampling periods, as it corresponds well with the temperature dependence of the vapor pressure. Considering the temperature difference between spring 2011 and 2013, the expected decrease in vapor pressure of iPCBs (ΔHvap: 80 000 J· mol−1) is 26%. This corresponds well with the 28% lower baseline concentration in 2013 compared to 2011 and supports our assumption that volatilization of PCBs is the major emission pathway. At the same time, it is also clear that, considering this temperature correction, we found no evidence for a general decrease in atmospheric PCB concentrations between 2011 and 2013. The rain fall rate was 1.0 and 4.6 mm in 2011 and 2013, respectively. As PCBs are known to be mostly in the gas phase at these environmental temperatures11,12,24 the rainfall induced wash-out of particles is not expected to affect iPCB concentrations in air substantially. C


Article


Figure 2. Spatial distribution of atmospheric iPCB concentrations in Zurich, Switzerland. The area of the blue circles represents the average of the iPCB concentrations measured in spring 2011 and 2013. This figure provides a comparison of iPCBs with concentrations of well-known air contaminants such as NO2 (a) and PM10 (b) and further parameters such as the population density (c) and the year of construction of all buildings in this urban region (d). The green-shaded area in (d) represents the district where approximately 40% of the buildings were constructed between 1950 and 1980. (Thematic maps are based on geodata from GIS-ZH. Maps reproduced with permission. Copyright 2014 swisstopo JD100043).

more detail below. Site C8 on the other hand was installed on the roof of a factory that produced electric capacitors and transformers since the 1930s. Prior to its ban, PCBs were commonly used as dielectric fluid in these electric products. Hence, a PCB contamination through spillage or volatile release is presumable at this location. Generally, elevated iPCB concentrations in air were predominately observed at the west end of Zurich. In contrast to this finding, the areas that experienced highest concentrations of conventional urban air pollutants such as NO2 or particulate matter (PM10) (Figure 2a,b) showed a different spatial distribution, as they were primarily located in the city center and along the main roads. PM10 emissions in urban areas are mainly caused by energy production through combustion, industrial activities, and road traffic,27 whereas the most important source of NO2 are motor vehicle emissions.28 The fundamental difference between the spatial distribution of these conventional urban air pollutants and iPCBs implies that combustion processes, industry, and traffic are minor emission pathways for PCBs in Zurich. In contrast to studies conducted in North America that observed maximum concentrations in the densely populated city center,10,12 atmospheric PCB concentrations in the center of Zurich were

Even though, it is commonly known that highly urbanized regions exhibit elevated PCB concentrations in air, the factors that cause the spatial variability of PCB concentrations within urban areas are unclear. It has been demonstrated that population density, road length, median household income, and proximity to contaminated sites are not able to fully explain the total variability.7,10 A case study in Toronto using land-use regression suggested that 75% of the PCB concentration variability was related to the distribution of PCBs in use, storage, and building sealants.26 Because these studies covered areas of several hundred square kilometers, the influence of small emission sources on air concentrations in the surrounding area could not be determined. In contrast, our study operates on a much smaller scale as 23 PAS were deployed in an urban area of around 50 km2. Although, most of the iPCB levels measured in Zurich were in the same range, concentrations at the sampling sites C4 and C8 were significantly higher than 1.5 times the interquartile range in both years (Figure 1). Therefore, we hypothesized that PCB-emitting sources have to be present in the surrounding area of these locations. Further investigations revealed that a housing complex with PCB-containing joint sealants is located close to the sampling site C4. This situation will be explained in D


Article


Figure 3. Coefficients of determinations for the linear correlation of measured iPCB concentrations in air with the number of buildings that were constructed between 1955 and 1975 in a perimeter of 100 to 500 m around all sampling sites in Zurich (a). Linear regression of the iPCB concentrations in air with the number of buildings that were constructed between 1955 and 1975 in a radius of 180 m, for which the highest coefficient of determination was observed (b).

comparably low. In general, PCB variability within the city of Zurich was not related to the population density (Figure 2c). Interestingly, the sampling sites with highest PCB concentrations in air are all situated in the same urban district (greenshaded area in Figure 2d). This urban district is apart from the historical center of Zurich and is characterized by a substantial population growth in the 1960s and 1970s. This effect is reflected in the construction activity, as approximately 40% of the buildings in this district were constructed between 1950 and 1980, whereas in the other urban districts of the study region maximum 30% of the building originated from this time period. This observation is crucial as the period of intense construction activity in this district corresponds well with the peak in global PCB production and usage.29 It is known that large volumes of PCBs were used in building materials.30 One major use was in joint sealants in order to increase their resistance to erosion, and to improve their flexibility.1 High concentrations of several percent of PCB by weight were frequently found in joint sealants of building constructed between 1950 and 1980.16,17,31 These sealants are of particular concern because they are applied to the outside of building and usually have a long lifecycle. Therefore, they potentially act as PCB sources over time periods of several decades. Relevance of Joint Sealants as PCB Sources. In Switzerland, large efforts were made to phase out PCBs in closed applications such as transformers and capacitors. Thereby, the inventory representing PCBs in these closed applications is estimated to have been lowered to less than 10 t in 2005.16 We therefore expect that PCBs in open systems such as joint sealants and paint are of major importance today. A previous study in Switzerland showed that one-third of the joint sealants in buildings constructed between 1966 and 1971 contained more than 10 g/kg of PCB.16 This finding is in accordance with the observation that emissions in the city of Zurich are primarily due to volatilization.15

It has been reported that PCB-containing joint sealants are associated with higher PCB concentrations in surrounding soils, as well as in indoor air and dust.30,32−34 However, this is the first study that focuses on the effect of contaminated joint sealants on urban ambient air. In Switzerland, polysulfide-based elastic joint sealants containing significant amounts of PCBs were mainly installed between 1955 and 1975. To investigate the correlation between measured iPCB concentrations in air and the occurrence of potentially PCB-contaminated joint sealants, we marked all buildings that were constructed in this time period with a red dot (Figure 2d). Generally, it seems that most of the locations that exhibited elevated PCB levels are surrounded by a high number of buildings that were constructed in the years 1955 to 1975. To assess this correlation quantitatively, we developed a mathematical approach. In a first step, we identified the total number of buildings that were constructed between 1955 and 1975 in a defined perimeter around every sampling site. In a second step, this number was then plotted against the average of the atmospheric iPCB concentration measured in 2011 and 2013 at the urban sampling site to investigate possible correlations. Sampling sites A5, A7, and B7 were excluded from this analysis as they were either not located in the city zone, or installed in a recreational area. Additionally, the site C8 was excluded, as this sampler was installed on the roof of a factory that produced electric equipment containing PCBs and, therefore, elevated concentrations may be due to a contaminated site. The size of the perimeter in this model approach is modifiable and was varied in the range of 100 to 500 m, to establish the best correlation. The lower boundary was set to 100 m to guarantee a substantial number of buildings around all sampling sites. We observed that the linear correlation was strongly depending on the radius of the specific perimeters (Figure 3). The highest coefficient of determination was achieved at a radius of 180 m (R2 = 0.75, p-value = 0.0023). E


Article

Environmental Science & Technology Smaller and bigger perimeters resulted in poorer correlation. The coefficients of determination at 100 and 500 m, for example, were smaller than 0.4, however, the linear regression was still significant at the 95% confidence level. Hence, we hypothesize that PCB-containing building material that was extensively used in this time period may influence air concentration in the surrounding area. This finding is supported by our previous study that concluded that PCB levels in Zurich may be caused by a large number of small and localized sources.15 For further investigation, data such as the type and cubic content of buildings, and information on renovation works, would ideally be known. Unfortunately this information was not available in the database. To evaluate the impact of PCBcontaining joint sealants on the surrounding air, we further assessed the emissions from a single source. Investigation of an Individual PCB Source. The sampling site C4 exhibited elevated iPCB concentrations in both years. An inspection of the environment around this site revealed a nearby housing complex with PCB-containing joint sealants. The measured iPCB content of these joint sealants was 20 g/kg (Figure S4, Supporting Information). The total PCB content can be estimated by multiplying the sum of iPCBs with a standard factor of 5,16 resulting in approximately 100 g/ kg. Commonly, joint sealants contained mostly mediumchlorinated PCBs, however, a relative depletion of the more volatile lower chlorinated PCBs over the years is likely.16,17 As the distance between this housing complex and the sampling site C4 is only about 120 m, we hypothesized that the high PCB concentrations measured at this site are caused by emissions from joint sealants attached to this building. To support this hypothesis and to preclude the presence of other PCB sources in the proximity of the site C4, 10 PAS air samplers were installed in this area (Figure 4). Thereby, the PAS at S5 corresponds to the site C4 as it was installed on the same rooftop. iPCB concentrations in air ranged from 125 pg· m−3 to 2676 pg·m−3 with highest levels next to the housing complex. The sites located in the east of the study area exhibit concentrations that are significantly higher than the urban baseline, whereas concentrations at sites in the west were lower and tend to approach the baseline much faster. Overall, iPCB concentrations are decreasing exponentially with increasing distance from the source and reach the baseline concentration in a distance of 150−300 m according to their spatial orientation. This distance corresponds well with our observation that iPCB levels at an individual sampling site are correlated with the number of buildings constructed between 1955 and 1975 in a radius of 150−250 m. Thus, it supports the hypothesis that PCB-containing building materials are important sources but only influence atmospheric concentrations in a spatially limited perimeter. Overall, the measured air concentrations were dominated by the low-chlorinated congeners at all sampling sites (Figure S5, Supporting Information), however, the proportion of highchlorinated congeners in the PCB pattern was increasing with growing distance from the housing complex. PCB-28 and PCB52 accounted for approximately 90% of total iPCBs at sites S1 and S2, whereas at sites S7 and S9 their portion was around 70%. This implies that air concentrations measured at sites next to the source are strongly affected by the volatilization of lowchlorinated PCB congeners from the joint sealants. Distant sites, on the other hand, are more similar to the congener patterns that were observed at sites that exhibit concentration

Figure 4. Map of the study area, where the buildings with PCBcontaining joint sealants are colored in red. Atmospheric levels of iPCBs are represented by yellow circles with areas that correspond to the concentrations observed at the sampling sites S1−S10. Absolute values of these measurement are indicated in in pg·m−3.

close to the urban baseline. All these findings support the hypothesis that the elevated PCB levels at the site C4 are primarily caused by emissions from this housing complex. Gaussian Plume Modeling. We developed a simple Gaussian plume model to explain the spatial distribution of atmospheric PCB concentration at this site and to extrapolate the source strength. These models are commonly used to determine the dispersion of nonreactive gaseous pollutants. For a continuous-point source of a pollutant, the Gaussian plume formula is expressed by eq 1:35 Cair(x , y , z) =

⎛ y 2 ⎞⎡ ⎛ (z − h)2 ⎞ Q ⎟ exp⎜⎜ − 2 ⎟⎟⎢exp⎜ − 2πμσyσz 2σz2 ⎠ ⎝ 2σy ⎠⎢⎣ ⎝ ⎛ (z + h)2 ⎞⎤ ⎟⎥ + exp⎜ − 2σz2 ⎠⎥⎦ ⎝ (1)

Where Cair (x,y,z) is the steady-state air concentration of a pollutant at the point (x,y,z) in g·m−3, x, y, and z are the coordinates for the downwind, crosswind, and vertical direction (m); Q is the source strength (g·s−1); u is the average wind speed (m·s−1), z and h are the heights of the source and the samplers at the sites S1−S10 (m); and σy and σz are the dimensionless lateral and vertical diffusion coefficients, respectively. F


Article


in the study area. However, the optimized total emission rates differed between the two scenarios and resulted in 6 μg·s−1 for the first set of meteorological parameters and 3.5 μg·s−1 for the second set. It is interesting to note that model results match well with measurements for the eastern sites (S5, S6, S8, S10) while the model output overestimates the concentrations at the western sites (S3, S4, S7, S9). This may be due to the inaccurate description of the local wind situation using meteorological data from distant stations. It is, for example, possible that the buildings of this housing complex disrupt the wind flow in northern direction which results in lower iPCB levels at sampling site in this direction. Based on the model results, extrapolated emissions from this housing complex were in the range of 110−190 g of iPCB per year. However, there remains considerable uncertainty regarding the simplicity of our model approach and the complexity of the wind situation in densely built-up areas. Although this housing complex is a big structure with 130 apartments, its emissions are still around a factor of 100 times lower than the total atmospheric iPCB emission of 13 kg per year in the city of Zurich that were estimated in our previous study.15 Thus, we can suppose that total emissions in Zurich are rather emitted from a larger number of small and localized sources than from a smaller number of large sources. Our study challenges the widespread opinion that PCBs represent legacy pollutants regulated decades ago that continuously lose their environmental and human health relevance. The appreciable amounts released by PCBcontaining building materials, point out that more efforts are required to further eliminate the existing sources. The sound characterization of the emissions of PCBs from a typical application such as building sealants we present here provides a significant contribution toward an efficient management of PCB inventories and development of corresponding PCB elimination strategies.37

The diffusion coefficients are a function of the downwind distance and depend on the atmospheric stability. Usually, Gaussian diffusion models are used with point sources; however, in this case the emissions appear to originate mainly from a housing complex consisting of several buildings that are partially connected to each other. We therefore modeled these buildings as volume sources according to the industrial source complex (ISC3) model that has been recommended by the USEPA as an air dispersion model for many years.36 The contribution of each individual building to the source strength of the total housing complex was calculated by dividing its cubic volume by the total volume of this housing complex. This model was used to calculate atmospheric iPCB concentrations at the sites S1−S10 on an hourly basis. The required input data including wind direction, wind speed, and the Pasquill Gifford stability class were derived from nearby meteorological stations (see Supporting Information). The purpose of this plume model was to reproduce the measured iPCB concentrations and to estimate the emission rate. As measurements represent mean concentration in air over the deployment time of the PAS, the model results were averaged over the corresponding time periods. In contrast to most of the cases where plume modeling is performed, the emission rate (Q) of the source is unknown. Therefore, it was determined iteratively with variable emissions until measured and modeled iPCB exhibited least absolute deviations. Further details and complete model equations can be found in the Supporting Information. The hourly resolved input data were collected from two meteorological stations that are located in the city of Zurich. The first one (M1) is located 4.5 km to the northeast, whereas the second one (M2) is located 2 km to the southwest of the study area. Analysis of the meteorological data revealed that the atmospheric stability was equivalent in most cases. The comparison of the wind roses for these sites (Figure S6 and S7), however, shows that wind directions and wind speed are quite different. Hence, we separately modeled the plume distribution with input parameters from the first and the second meteorological station individually (Figure 5). Generally, plume modeling with both data sets is able to reproduce most of the spatial variation of iPCB concentrations

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b04626. Additional text, tables and figures as mentioned in the text(PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +41 58 765 4953. Fax: +41 58 765 6963. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are thankful to Athanasios Katsoyiannis (Lancaster University, now JRC Ispra) for providing the sampling equipment. The Swiss National Air Pollution Monitoring Network (NABEL) is acknowledged for providing access to their sampling sites and the Federal Office for Meteorology and Climatology (MeteoSwiss) for providing the meteorological data. The authors are also grateful for access to the numerous public and private sampling sites within the city of Zurich. The authors thank the Swiss Federal Office of the Environment (FOEN) and the Office of Waste, Water, Energy and Air of the canton of Zurich (WWEA) for financial support.

Figure 5. Comparison of measured and modeled iPCB concentrations in air at sampling sites with a distance of 50 to 300 m from a housing complex with PCB-containing joint sealants. Black bars represent the PAS measurements, and white and gray bars are the results of the Gaussian plume model calculation with input data from two different meteorological stations (M1 and M2). G


Article


■

(19) Bohlin, P.; Audy, O.; Skrdlíková, L.; Kukučka, P.; Přibylová, P.; Prokeš, R.; Vojta, S.; Klánová, J. Outdoor passive air monitoring of semi volatile organic compounds (SVOCs): a critical evaluation of performance and limitations of polyurethane foam (PUF) disks. Environ. Sci. Process. Impacts 2014, 16 (3), 433−444. (20) Mari, M.; Schuhmacher, M.; Feliubadaló, J.; Domingo, J. L. Air concentrations of PCDD/Fs, PCBs and PCNs using active and passive air samplers. Chemosphere 2008, 70 (9), 1637−1643. (21) Diefenbacher, P. S.; Bogdal, C.; Gerecke, A. C.; Glüge, J.; Schmid, P.; Scheringer, M.; Hungerbühler, K. Short-Chain Chlorinated Paraffins in Zurich, SwitzerlandAtmospheric Concentrations and Emissions. Environ. Sci. Technol. 2015, 49 (16), 9778−9786. (22) Menichini, E.; Iacovella, N.; Monfredini, F.; Turrio-Baldassarri, L. Atmospheric pollution by PAHs, PCDD/Fs and PCBs simultaneously collected at a regional background site in central Italy and at an urban site in Rome. Chemosphere 2007, 69 (3), 422−434. (23) Cetin, B.; Yatkin, S.; Bayram, A.; Odabasi, M. Ambient concentrations and source apportionment of PCBs and trace elements around an industrial area in Izmir, Turkey. Chemosphere 2007, 69 (8), 1267−1277. (24) Mandalakis, M.; Stephanou, E. G. Atmospheric concentration characteristics and gas-particle partitioning of PCBs in a rural area of eastern Germany. Environ. Pollut. 2007, 147 (1), 211−221. (25) Jamshidi, A.; Hunter, S.; Hazrati, S.; Harrad, S. Concentrations and chiral signatures of polychlorinated biphenyls in outdoor and indoor air and soil in a major U.K. conurbation. Environ. Sci. Technol. 2007, 41 (7), 2153−2158. (26) Melymuk, L.; Robson, M.; Helm, P. A.; Diamond, M. L. Application of land use regression to identify sources and assess spatial variation in urban SVOC concentrations. Environ. Sci. Technol. 2013, 47 (4), 1887−1895. (27) Keuken, M. P.; Moerman, M.; Voogt, M.; Blom, M.; Weijers, E. P.; Röckmann, T.; Dusek, U. Source contributions to PM2.5 and PM10 at an urban background and a street location. Atmos. Environ. 2013, 71 (x), 26−35. (28) Mueller, M. D.; Wagner, M.; Barmpadimos, I.; Hueglin, C. Twoweek NO2 maps for the City of Zurich, Switzerland, derived by statistical modelling utilizing data from a routine passive diffusion sampler network. Atmos. Environ. 2015, 106 (2), 1−10. (29) Breivik, K.; Sweetman, A.; Pacyna, J. M.; Jones, K. C. Towards a global historical emission inventory for selected PCB congeners–a mass balance approach. 1. Global production and consumption. Sci. Total Environ. 2002, 290, 181−198. (30) Andersson, M.; Ottesen, R. T.; Volden, T. Building materials as a source of PCB pollution in Bergen, Norway. Sci. Total Environ. 2004, 325 (1−3), 139−144. (31) Klosterhaus, S.; McKee, L. J.; Yee, D.; Kass, J. M.; Wong, A. Polychlorinated biphenyls in the exterior caulk of san francisco bay area buildings, california, USA. Environ. Int. 2014, 66, 38−43. (32) Priha, E.; Hellman, S.; Sorvari, J. PCB contamination from polysulphide sealants in residential areas - Exposure and risk assessment. Chemosphere 2005, 59 (4), 537−543. (33) Herrick, R. F.; McClean, M. D.; Meeker, J. D.; Baxter, L. K.; Weymouth, G. A. An unrecognized source of PCB contamination in schools and other buildings. Environ. Health Perspect. 2004, 112 (10), 1051−1053. (34) Kohler, M.; Zennegg, M.; Waeber, R. Coplanar polychlorinated biphenyls (PCB) in indoor air. Environ. Sci. Technol. 2002, 36 (22), 4735−4740. (35) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; 2006; Vol. 2. (36) U.S. EPA. User’s guide for the industrial source complex (ISC3) dispersion models, Volume II - Description of model algorithms. In EPA-454/B-95−003a; Research Triangle Park, NC, 1995; pp 1−128. (37) Diamond, M. L.; Melymuk, L.; Csiszar, S. A.; Robson, M. Estimation of PCB stocks, emissions, and urban fate: will our policies reduce concentrations and exposure? Environ. Sci. Technol. 2010, 44 (8), 2777−2783.

REFERENCES

(1) Erickson, M. D.; Kaley, R. G. Applications of polychlorinated biphenyls. Environ. Sci. Pollut. Res. 2011, 18 (2), 135−151. (2) Totten, L. A.; Gigliotti, C. L.; Vanry, D. A.; Offenberg, J. H.; Nelson, E. D.; Dachs, J.; Reinfelder, J. R.; Eisenreich, S. J. Atmospheric Concentrations and Deposition of Polychorinated Biphenyls to the Hudson River Estuary. Environ. Sci. Technol. 2004, 38 (9), 2568−2573. (3) Harner, T.; Shoeib, M.; Diamond, M.; Stern, G.; Rosenberg, B. Using passive air samplers to assess urban-rural trends for persistent organic pollutants. 1. Polychlorinated biphenyls and organochlorine pesticides. Environ. Sci. Technol. 2004, 38 (17), 4474−4483. (4) Motelay-Massei, A.; Harner, T.; Shoeib, M.; Diamond, M.; Stern, G.; Rosenberg, B. Using passive air samplers to assess urban-rural trends for persistent organic pollutants and polycyclic aromatic hydrocarbons. 2. Seasonal trends for PAHs, PCBs, and organochlorine pesticides. Environ. Sci. Technol. 2005, 39 (15), 5763−5773. (5) Gingrich, S. E.; Diamond, M. L.; Stern, G. A.; McCarry, B. E. Atmospherically derived organic surface films along an urban-rural gradient. Environ. Sci. Technol. 2001, 35 (20), 4031−4037. (6) Wethington, D. M.; Hornbuckle, K. C. Milwaukee, WI, as a source of atmospheric PCBs to Lake Michigan. Environ. Sci. Technol. 2005, 39 (1), 57−63. (7) Du, S.; Wall, S. J.; Cacia, D.; Rodenburg, L. A. Passive air sampling for polychlorinated biphenyls in the Philadelphia metropolitan area. Environ. Sci. Technol. 2009, 43 (5), 1287−1292. (8) Persoon, C.; Peters, T. M.; Kumar, N.; Hornbuckle, K. C. Spatial distribution of airborne polychlorinated biphenyls in Cleveland, Ohio and Chicago, Illinois. Environ. Sci. Technol. 2010, 44 (8), 2797−2802. (9) Hsu, Y. K.; Holsen, T. M.; Hopke, P. K. Locating and quantifying PCB sources in Chicago: Receptor modeling and field sampling. Environ. Sci. Technol. 2003, 37 (4), 681−690. (10) Hu, D.; Lehmler, H.-J.; Martinez, A.; Wang, K.; Hornbuckle, K. C. Atmospheric PCB congeners across Chicago. Atmos. Environ. 2010, 44 (12), 1550−1557. (11) Kuzu, S. L.; Saral, A.; Summak, G.; Coltu, H.; Demir, S. Ambient polychlorinated biphenyl levels and their evaluation in a metropolitan city. Sci. Total Environ. 2014, 472, 13−19. (12) Melymuk, L.; Robson, M.; Helm, P. A.; Diamond, M. L. PCBs, PBDEs, and PAHs in Toronto air: Spatial and seasonal trends and implications for contaminant transport. Sci. Total Environ. 2012, 429, 272−280. (13) Gasic, B.; Moeckel, C.; MacLeod, M.; Brunner, J.; Scheringer, M.; Jones, K. C.; Hungerbühler, K. Measuring and modeling shortterm variability of PCBs in air and characterization of urban source strength in Zurich, Switzerland. Environ. Sci. Technol. 2009, 43 (3), 769−776. (14) Bogdal, C.; Müller, C. E.; Buser, A. M.; Wang, Z.; Scheringer, M.; Gerecke, A. C.; Schmid, P.; Zennegg, M.; Macleod, M.; Hungerbühler, K. Emissions of polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins, and polychlorinated dibenzofurans during 2010 and 2011 in Zurich, Switzerland. Environ. Sci. Technol. 2014, 48 (1), 482−490. (15) Diefenbacher, P. S.; Bogdal, C.; Gerecke, A. C.; Glüge, J.; Schmid, P.; Scheringer, M.; Hungerbü hler, K. Emissions of Polychlorinated Biphenyls in Switzerland: A Combination of LongTerm Measurements and Modeling. Environ. Sci. Technol. 2015, 49 (4), 2199−2206. (16) Kohler, M.; Tremp, J.; Zennegg, M.; Seiler, C.; Minder-Kohler, S.; Beck, M.; Lienemann, P.; Wegmann, L.; Schmid, P. Joint sealants: an overlooked diffuse source of polychlorinated biphenyls in buildings. Environ. Sci. Technol. 2005, 39 (7), 1967−1973. (17) Robson, M.; Melymuk, L.; Csiszar, S. A.; Giang, A.; Diamond, M. L.; Helm, P. A. Continuing sources of PCBs: the significance of building sealants. Environ. Int. 2010, 36 (6), 506−513. (18) Chaemfa, C.; Barber, J. L.; Gocht, T.; Harner, T.; Holoubek, I.; Klanova, J.; Jones, K. C. Field calibration of polyurethane foam (PUF) disk passive air samplers for PCBs and OC pesticides. Environ. Pollut. 2008, 156 (3), 1290−1297. H


Spatial Distribution of Atmospheric PCBs in Zurich, Switzerland: Do

Recommend Documents