Environ. Sci. Technol. 2005, 39, 4721-4728
Environmental and Human Impact of an Old-Timer Incinerator in Terms of Dioxin and PCB Level: A Case Study CATHERINE PIRARD,* GAUTHIER EPPE, A N N E - C EÄ C I L E M A S S A R T , S EÄ B A S T I E N F I E R E N S , † EDWIN DE PAUW, AND J E A N - F R A N C¸ O I S F O C A N T Mass Spectrometry Laboratory, C.A.R.T., University of Lie`ge, Alle´e de la chimie, 3, Baˆt B6c, B-4000 Lie`ge, Belgium
The impact of a recently closed old municipal solid waste incinerator (MSWI) on polychlorodibenzo-p-dioxin (PCDD), polychlorodibenzofuran (PCDF), and polychlorinated biphenyl (PCB) levels in the surrounding environment and resident serum has been studied in a small rural area of France. Studied soils and eggs from chickens foraging on these soils were sampled in the vicinity of the MSWI under the prevailing wind stream, while comparison samples were collected outside the assumed impact zone. PCB levels observed in soils and eggs did not differ statistically from comparison sites. This confirmed the low impact of MSWI PCB emission on environmental media, compared to other well-known sources. PCDD/PCDF levels in soils and eggs were significantly higher than in comparison samples, pointing out the impact of MSWI emission on the surrounding environment. The high dioxin concentrations in eggs set aside for private consumption would increase the dioxin intake for the studied population. Blood specimens of 10 nonoccupationally exposed volunteers who had lived within a 2 km radius of the incinerator for at least 25 years have been analyzed. When adjusted for age, PCB and PCDD/F blood levels were higher than general European populations and comparable to a similarly exposed Belgian population.
Introduction Polychlorodibenzo-p-dioxins (PCDDs) and polychlorodibenzofurans (PCDFs) are unintentionally produced molecules that were originally identified (1) in fly ash and flue gas of some municipal incinerators in 1977. Since then, this class of pollutants has been found at trace levels in the environment and in abiotic samples as well as all along the trophic pyramid. Several industrial sources contribute to dioxin emissions in the atmosphere; the most important are municipal solid waste incinerators (MSWIs) and metal industry (2, 3). Because very strict controls are currently applied in the industrial sector, in particular to MSWIs, diffuse sources such as traffic, accidental fires, domestic coal combustion, and backyard burning of domestic waste * Corresponding author phone: +32 (0)4 366 3531; fax: +32 (0)4 366 4387; e-mail:
[email protected]. † Current address: Industrial Toxicology and Occupational Medicine Unit, Catholic University of Louvain, Brussels, Belgium. 10.1021/es0481981 CCC: $30.25 Published on Web 05/25/2005
2005 American Chemical Society
increasingly contribute to the total dioxin emissions (2, 4). In France, the European directive (EC94/67/CE) fixing the threshold PCDD/F level at 0.1 ng TEQ/Nm3 has been applied to new MSWIs having a capacity above 6 tons/h since 1997 (5). As a result of these stringent regulations, several old MSWIs that did not comply with legal norms have been closed over the last 5 years. Despite this, it is assumed that large amounts of PCDD/Fs have potentially been released into the environment prior MSWIs closure. The PCB question is completely different since they were intentionally manufactured for closed (i.e. transformers and capacitors) or open applications (i.e. adhesive, plasticizer, ink, paint). Consequently, these past uses and their illegal disposal are one of the major sources of PCBs in the environment in addition to some chemical processes (6). However, it has been found that PCBs are also emitted from industrial sources, even if they weakly contribute to total TEQ (6, 7). Because soils have been demonstrated to act as a conservative matrix for dioxin or PCB atmospheric deposition and to reflect a long-term exposure (8), the impact of pollution from these old factories would still be perceptible in the surrounding grounds (9, 10). Foraging animals such as free-range chickens are particularly exposed to soil contamination by direct soil ingestion. They are known to accumulate persistent pollutants and transfer them to their eggs (11-13). It is reasonable to assume that people consuming these contaminated food products would increase their dioxin and PCB body burden, resulting in an elevation of their serum levels. Unfortunately, correlation between dioxin exposure and blood levels is not so clear, especially as many factors (such as age, gender, smoking, and eating habits, etc.) can affect serum levels. Wide variations usually occur across populations, often making any comparison difficult. This study provided an overall view of the impact of an old French incinerator on dioxin and PCB levels in soil, in locally produced animal food and in resident serum. Sites located less than 2 km away from MSWI under the prevailing wind stream were compared to areas situated outside the assumed potential impact zone. To our knowledge this is the first report on dioxin blood levels in the French population.
Materials and Methods Study Area. This study was conducted in Maincy (France), a rural area with a population of 1700 located 60 km south of Paris and 5 km west of Melun, an industrial city with 40 000 inhabitants. Less than 2 km away from Maincy stands the Vaux-le-Pe´nil MSWI. It began operating in 1974 and was closed in June 2002 following a dioxin measures campaign for atmospheric emissions. The facility, with a capacity of 4 metric tons per hour, handled approximately 40 000 metric tons of domestic waste per year, without any specific dioxin removal. Measured dioxin emission level was 226 ng TEQ/ Nm3, more than 2000-fold the current European norm set at 0.1 ng TEQ/Nm3 (5). At the same time, soil and plant samples close to the incinerator were found to have similar dioxin levels to those in other urban or industrial areas, while dioxin levels in milk from cows grazing in the vicinity was calculated at around 9 pg TEQ/g of fat. There was also a clinical waste incinerator (CWI) located in Melun, which opened in 1988, handling 0.5 t/h. Dioxin emissions were recorded at 3.9 ng TEQ/m3 when it closed in 2000. According to the wind statement, the village of Maincy was not in the direction of the prevailing wind from the CWI. Sampling. A number of sites were selected in Maincy (study area) for soil and egg sampling. These were located VOL. 39, NO. 13, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4721
TABLE 1. Characteristics of Sites Selected for Specimen Collection site
distance from MSWI (m)
1 2 3 4 5 6 7 8 9 10
1500 1250 1250 1500 1000 1000 2250 1300 1250 1500
NE NE NE NE NE NE NE NE NE N
11 12 13 14 15 16
1500 8500 1500 8500 4700 5300
S NE N-NW NE N-NW N-NE
direction
type of housing Study Area (Maincy) lawn agric land henhouse henhouse lawn lawn lawn lawn lawn lawn Comparison Area henhouse lawn henhouse henhouse lawn lawn
between 1 and 1.5 km from the MSWI, under the prevailing wind stream (Northeast). Sites used as comparison areas were assumed to be outside the potential impact zone and were situated either under nondominant wind stream or in the Northeast direction but at 8.5 km away from the incinerator. Table 1 shows characteristics of selected areas. For each site, soil was sampled 10 cm deep, at 2 distinct randomly selected spots in private gardens where chickens were foraging. Samples were oven-dried overnight and then stored in polyethylene (PE) sampling bags at room temperature. For each site 4-6 eggs were hard-boiled. Yolks were isolated and pooled prior to storage in PE vials at -20 °C. Blood samples (100 mL) were taken from 10 volunteers who had lived within a 2 km radius of the incinerator for at least 25 years. Samples were collected between February and April 2003. The group of participants included 2 females and 8 males between the ages of 51 and 75 years (mean of 61 years old), with no reported occupational exposure in the past. They were asked to complete a questionnaire that included data on age, height, body weight, smoking habits, and frequency of consumption of eggs from locally raised chickens as well as vegetables grown in their own gardens. Serum was separated from clotted blood by centrifugation and stored at -20 °C prior to analyses. Analytical Method. Chemicals, standard solutions, and detailed procedures for the analysis of PCDD/Fs and PCBs in serum and eggs have previously been extensively described by Focant et al. (14, 15) and Pirard et al. (16). Briefly, formic acid and water were added to serum prior to extraction by solid-phase extraction (SPE) using hexane to elute selectively organic compounds from the C18-linked silica core. Extraction from eggs and soils was performed by pressurized liquid extraction (ASE 200, Dionex, Sunnyvale, CA). Hexane was used for hard-boiled yolks mixed with sodium sulfate, and toluene was used for soils to which acidic silica gel had been added prior to blending. Further cleanup of all extracts was carried out with the automated Power-Prep (FMS Inc., Waltham, MA) system using acidic silica, basic alumina, and PX-21 carbon columns (16) and an additional high-capacity disposable silica column for yolk extract (14). The PCB fraction, containing 7 indicator and 8 monoortho PCBs, was analyzed by tandem in time mass spectrometry (GC-MS/MS) using a PolarisQ ion trap mass spectrometer ThermoQuest (Austin, TX, U.S.A.) coupled with a gas chromatograph. Final measurement of PCDD/Fs and coplanar PCBs (c-PCBs) was performed by GC-HRMS using an Autospec Ultima (Micromass, Manchester, U.K.) high4722
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 13, 2005
household waste fire
major heating type
chimney wood fire
no no
natural gas
yes
yes no yes yes yes yes
wood stove fuel oil fuel oil natural gas wood stove electricity fuel oil electricity
yes no yes yes yes no no yes
no yes no no no yes
fuel oil natural gas wood stove natural gas fuel oil electricity
yes no yes no yes yes
resolution mass spectrometer and an Agilent (Palo Alto, CA) 6890 Series GC equipped with an Rtx 5-MS (40 m × 0.18 mm i.d. × 0.20 µm film thickness) capillary column (Restek, Interscience, Belgium). Details concerning operating conditions, tuning, and checks to ensure operation of the MS system have been previously reported (14, 16). Total serum lipids were estimated by an enzymatic method measuring triglycerids, phospholipids, cholesterol, and free cholesterol (17), using commercially available reagents from Wako Chemicals (Neuss, Germany). Quality Assurance. Confidence in the methodology for the determination of PCDD/Fs and c-PCBs in eggs and serum is provided by accreditation according to the ISO 17025 standard. An analytical method for dioxin analysis in soils was evaluated in several interlaboratory round tests, and PCB results on serum samples were successfully compared with those obtained by another accredited laboratory. Data Analysis. Toxic equivalence to 2,3,7,8-TCDD (TEQ) were evaluated according to toxic equivalency factors (TEF) defined by the World Health Organization (WHO) (18). Measurements below limits of quantification (LOQs) were treated as half the LOQ value of the congener considered. Analysis of variance (Anova) and Principal Components Analysis (PCA) were calculated using Minitab 10.5 Xtra statistical software. ANOVA was used to compare pollutant mean levels in soils and eggs from Maincy and soils and eggs from comparison sites. Since distributions deviated significantly from normality (i.e. PCDD/Fs levels in soils and eggs), thus the nonparametric Mann-Withney test was applied to test the differences between levels from Maincy and comparison areas. For both the study and the comparison areas, the sum of indicator PCBs did not include PCB-28 and -52 because of their high LOQ values at the time of the measurement.
Results Concentrations of the sums of the 2,3,7,8-substituted dioxins and furans, the 4 non-ortho PCBs (IUPAC number: 77, 81, 126, 169), the 8 mono-ortho (105, 114, 118, 123, 156, 157, 167, 189) PCBs, and the 5 of the 7 indicator PCBs (101, 118, 138, 153, 180) are given in Table 2 for soils and eggs on each site from Maincy and from the comparison areas. Levels for PCDD/Fs, non- and mono-ortho PCBs were expressed in pg TEQ/g of dry matter (d.m.) for soils and pg TEQ/g fat for eggs, while indicator PCBs were evaluated in ng/g of d.m. and ng/g fat for soil for egg samples, respectively.
TABLE 2. Concentrations of PCBs and PCDD/Fs in Soil and Egg Samples from Each Site and Median Concentrations Calculated for All Comparison and Study Sitesd soils
eggs
PCBs
PCBs
site
indicatora
m-o
n-o
PCDD/Fs
1 2 3 4 5 6 7 8 9 10 xmedian
5.2 17.2
0.2 0.2
41.2 21.9 5.2 15.8 24.4
0.4 0.6 0.2 0.2 0.7
1.1 0.8 1.5 1.5 1.7 0.7 0.5 1.4
13.4 11.1 12.8 19.9 59.0 11.6 12.2 37.2
8.7 17.2
0.2 0.2
0.5 1.26
3.3 13.1*
11 12 13 14 15 16 xmedian
14.8 5.8 30.8 6.7 5.8 5.6 7.3
0.3 0.2 3.0 0.4 0.2 0.2 0.2
1.7 0.4 4.7 1.8 0.5 0.9 0.9
6.6 0.3 5.6 2.6 1.6 3.0 3.0*
a Concentration in ng/g dry matter. study and comparison areas.
b
total TEQc
indicatorb
Study Area (Maincy) 14.8 24.9 12.1 47.6 14.4 40.2 21.8 117.1 61.4 34.6 12.5 43.9 12.9 23.8 39.3 283.1 106.3 4.0 133.9 14.6* 49.8 Comparison Area 8.5 0.9 13.3 4.4 2.3 4.2 4.1*
Concentration in ng/g fat. c WHO-TEQ (18).
PCB Levels in Soils, Eggs, and Resident Serum. The pentachlorinated PCB-101 was the most abundant congener found in soils, with levels ranging from 1.9 to 26.3 ng/g d.m, followed by PCB-138, -153, and -180 with approximately similar levels. The pentachlorinated PCB-126 was also predominant for coplanar congeners, with amounts varying from 0.4 to 1.8 pg TEQ/g d.m., while PCB-169 was constantly inferior to 0.0014 pg TEQ/g, LOQ value for this matrix. MannWithney test results indicated no statistically significant differences between PCB levels in soils from Maincy and from comparison areas. The most important PCB congeners in eggs from all sites were PCB-138 and -153, with similar levels ranging from 4.3 to 98.6 ng/g fat and 2.8 to 105 ng/g fat, followed by PCB-118, with a median quantity of 11.1 ng/g fat. Other PCB congeners were less abundant. There were two spots (sites 8 and 11) where the sum of indicators in collected eggs exceeded the legal European norm of 200 ng/g fat (299 and 310 ng/g fat), but the others remained considerably below this threshold. PCB-126 accounted for 99% of the TEQ contribution from coplanar PCBs, representing nearly one-third of the total (PCDD/F and PCB) TEQ level. Median and range concentrations of individual indicator PCBs and non- and mono-ortho PCBs in serum samples are given in Table 3. Levels were expressed in pg TEQ/g lipid for dioxin-like PCBs and in ng/g lipid for indicator PCBs. PCB180, -153, and -138 were dominant, each accounting for 2530% of the total non-dioxin-like PCB burden. PCB-156 was the PCB congener that contributed the most to total TEQ, with a mean level around 12 pg TEQ/g lipid, followed by PCB-126, with about 9 pg TEQ/g lipid. In other non- or monoortho PCBs, only PCB-118, and to a lesser extent PCB-157 and -169, significantly increased the TEQ value. PCBs-114, -123, and -189 were not detected in any samples. PCDD/F Levels in Soils, Eggs, and Resident Serum. Contrary to those for the PCBs, Mann-Withney test results indicated statistically significant differences between PCDD/F levels in soils from Maincy and from comparison areas. Dioxin and furan patterns were similar, regardless of the sampling area or surface type (lawn, agricultural land, or upper manure layer from a henhouse). Octachlorodibenzo-p-dioxin (OCDD)
274.5 62.3 143.2 143.5 140.1 21.4 143.3 d
m-o
n-o
PCDD/Fs
total TEQc
0.7 1.4 1.1 3.8 1.0 1.0 0.5 10.5 3.0 5.6 1.1
11.7 23.6 8.8 19.6 9.9 4.6 0.3 16.2 6.1 13.4 9.9
71.4 121.6 24.3 95.4 86.2 6.3 5.1 25.7 14.5 10.7 25.7*
83.8 146.5 34.2 118.8 97.0 11.8 5.9 52.5 23.7 29.6 36.7*
6.1 2.1 7.6 4.0 4.6 0.5 4.6
46.4 9.9 23.7 11.8 8.3 3.1 11.8
4.4 7.7 8.1 6.7 11.2 3.1 7.7*
56.9 19.7 39.5 22.5 24.1 6.7 24.2*
Asterisks indicate statistically significant differences between
levels constituted between 50 and 90% of the total in pg/g d.m., except for two samples having 25-35% (sites 2 and 6). Other main congeners present in soils were 1,2,3,4,6,7,8heptachlorodibenzo-p-dioxin (HpCDD), accounting for about 10-20%, OCDF, and 1,2,3,4,6,7,8-heptachlorodibenzofuran (HpCDF 1), both for about 5%. PCA, a multivariate analysis technique allowing dimension reduction, was performed to compare pattern of soils from this study and some European data gathered from recent literature and detailed in Table 4. The 17 variables were concentrations of individual 2,3,7,8substituted congeners in terms of percentage of the sum of all 2,3,7,8-substituted compounds. The first three components would explain 79.8% of the variance of the data set. The first main principal component (PC1), accounting for 54.2% of the total variability, was highly positively correlated with 2,3,4,7,8-pentachlorodibenzofuran (PeCDF 2), 1,2,3,6,7,8hexachlorodibenzofuran (HxCDF 2), 2,3,4,6,7,8-hexachlorodibenzofuran (HxCDF 4), HpCDF 1, and low chlorinated dioxins (tetrachlorodibenzo-p-dioxin, pentachlorodibenzop-dioxin, and hexachlorodibenzo-p-dioxins) and negatively correlated with OCDD. PC2 retained 16.7% of the total variance and was positively correlated with HpCDD and negatively with high octachlorodibenzofuran (OCDF), 1,2,3,4,7,8-hexachlorodibenzofuran (HxCDF 1), 1,2,3,7,8,9hexachlorodibenzofuran (HxCDF 3), and 1,2,3,4,7,8-hexachlorodibenzofuran (PeCDF 1). Figure 1 shows the twodimensional principal components score plot of soils from our study and from other European areas. All egg samples contained dioxin and furan levels superior to the 3 pg TEQ/g fat maximum level applied in Europe. As for soils, the Mann-Withney test demonstrated statistical difference between PCDD/F levels in eggs from Maincy and from the comparison area but no difference for PCBs. There was no apparent relationship between dioxin and furan levels and PCB levels. Nevertheless, for most samples, high dioxin concentrations meant high PCB concentrations, except in 2 sites (6 and 11), which contained quite high intensities of PCB and low amounts of dioxin. Median and range concentrations of individual 2,3,7,8PCDD/Fs in serum samples are given in Table 3, with levels expressed in pg TEQ/g lipid. OCDD considerably dominated VOL. 39, NO. 13, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4723
TABLE 3. Median and Range of Concentrations for Indicator PCBs, Non- and Mono-Ortho PCBs, and PCDD/Fs in Serum Samples of Volunteers Living for at Least 25 Years within a 2 km Radius of the Incineratora congeners
median Indicator PCBs (ng/g lipid) 23 144 187 189 587
19-36 125-271 145-268 149-297 [377-656]
PCB-105 PCB-114 PCB-118 PCB-123 PCB-156 PCB-157 PCB-167 PCB-189 total m-o PCBs
m-o PCBs (pg TEQ/g lipid) 0.38 ND 5.46 ND 11.99 1.95 0.08 ND 17.91
0.1-1.3 ND 4.4-7.2 ND 9.3-23.6 0.7-4.1 0.07-0.1 ND [14.8-29.0]
PCB-77 PCB-126 PCB-169 total n-oPCBs
n-o PCBs (pg TEQ/g lipid)
