Deposition fluxes of PCDD/Fs in the area surrounding a steel plant in northwest Italy Maurizio Onofrio & Roberta Spataro & Serena Botta Received: 15 May 2013 /Accepted: 28 January 2014 # Springer International Publishing Switzerland 2014 Abstract The paper aims at investigating the contribu- tion of a steel plant located in a rural area in northwest- ern Italy (700,000 tons of steel/year) to the deposition fluxes of Polychorinated Dibenzo-p-dioxins and Polychorinated dibenzofurans (PCDD/PCDFs) at local level through the analysis of sampling data, literature data, and air dispersion model (AERMOD)output data. Total measured deposition fluxes of PCDD/PCDFs in three monitoring stations were consistent with other studies carried out in Italy in urban and suburban areas and in rural European areas; while these were lower than those measured in other European urban/suburban areas or in sites influenced by industrial sources. Furthermore, the measured fluxes were also compared with the pattern of PCDD/Fs in ambient air sampled at the same sites in a previous study. This comparison showed a similarity between air concentration and deposition patterns of the samples collected at the three monitoring stations and a clear distinction of these from the source. The study was completed with AERMOD simulations, conducted with a mass mean particle diameter of 0.5 μm, according to the particle size distribution of the samples collected at the source. AERMOD calculated deposition fluxes of two to three orders of magnitude lower than those measured in two monitoring points; while in the most distant monitor- ing station, the deposition fluxes were too low to be calculated by the model. The simulations confirmed that the most distant monitoring station was not subject to emissions from the steel plant. The analysis highlighted the limited influence of the source in the local PCDD/F deposition fluxes. Keywords PCDD/Fs . Air dispersion model . Deposition . Steel plant emissions . PCA Introduction Polychlorinated dibenzo- p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) are released in- to the atmosphere from combustion processes and dif- ferent types of industrial sources, mainly waste inciner- ators, steel mills, metal smelting and chlorination (Bakoglu et al. 2005; Colombo et al. 2009; Li et al. 2010; Mari et al. 2008), and traffic (Benfenati et al. 1992; Chetwittayachan et al. 2002). Air is the main distribution pathway for PCDD/Fs (Lohmann and Jones 1998), and atmospheric deposition is one of the most important processes for the transport of PCDD/Fs from numerous emission sources to the envi- ronmental compartments. In particular, the deposition of PCDD/Fs from the air to different matrices (soil, water, and plants) leads to their entry into the food chain (Lohmann and Jones 1998). PCDD/Fs are extremely re- sistant to both environmental and biological degradation, and they are able to bioaccumulate (Schecter et al. 2006). There are many available studies aimed at investigat- ing different aspects of PCDD/Fs deposition, such as Environ Monit Assess DOI 10.1007/s10661-014-3668-y M. Onofrio : R. Spataro : S. Botta (*) Department of Engineering of Environment, Land and Infrastructures, Politecnico di Torino, C.so Duca degli Abruzzi, 24, 10129 Turin, Italy e-mail: [email protected]
Transcript
Deposition fluxes of PCDD/Fs in the area surrounding a steelplant in northwest Italy
Maurizio Onofrio & Roberta Spataro & Serena Botta
Received: 15 May 2013 /Accepted: 28 January 2014# Springer International Publishing Switzerland 2014
Abstract The paper aims at investigating the contribu-tion of a steel plant located in a rural area in northwest-ern Italy (700,000 tons of steel/year) to the depositionfluxes of Polychorinated Dibenzo-p-dioxins andPolychorinated dibenzofurans (PCDD/PCDFs) at locallevel through the analysis of sampling data, literaturedata, and air dispersion model (AERMOD)output data.Total measured deposition fluxes of PCDD/PCDFs inthree monitoring stations were consistent with otherstudies carried out in Italy in urban and suburban areasand in rural European areas; while these were lower thanthose measured in other European urban/suburban areasor in sites influenced by industrial sources. Furthermore,the measured fluxes were also compared with the patternof PCDD/Fs in ambient air sampled at the same sites in aprevious study. This comparison showed a similaritybetween air concentration and deposition patterns of thesamples collected at the three monitoring stations and aclear distinction of these from the source. The study wascompleted with AERMOD simulations, conducted with amass mean particle diameter of 0.5 μm, according to theparticle size distribution of the samples collected at thesource. AERMOD calculated deposition fluxes of two tothree orders of magnitude lower than those measured intwo monitoring points; while in the most distant monitor-ing station, the deposition fluxes were too low to be
calculated by the model. The simulations confirmed thatthe most distant monitoring station was not subject toemissions from the steel plant. The analysis highlightedthe limited influence of the source in the local PCDD/Fdeposition fluxes.
Keywords PCDD/Fs . Air dispersionmodel .
Deposition . Steel plant emissions . PCA
Introduction
Polychlorinated dibenzo-p-dioxins (PCDDs) andpolychlorinated dibenzofurans (PCDFs) are released in-to the atmosphere from combustion processes and dif-ferent types of industrial sources, mainly waste inciner-ators, steel mills, metal smelting and chlorination(Bakoglu et al. 2005; Colombo et al. 2009; Li et al.2010; Mari et al. 2008), and traffic (Benfenati et al.1992; Chetwittayachan et al. 2002).
Air is the main distribution pathway for PCDD/Fs(Lohmann and Jones 1998), and atmospheric depositionis one of the most important processes for the transport ofPCDD/Fs from numerous emission sources to the envi-ronmental compartments. In particular, the deposition ofPCDD/Fs from the air to different matrices (soil, water,and plants) leads to their entry into the food chain(Lohmann and Jones 1998). PCDD/Fs are extremely re-sistant to both environmental and biological degradation,and they are able to bioaccumulate (Schecter et al. 2006).
There are many available studies aimed at investigat-ing different aspects of PCDD/Fs deposition, such as
Environ Monit AssessDOI 10.1007/s10661-014-3668-y
M. Onofrio :R. Spataro : S. Botta (*)Department of Engineering of Environment,Land and Infrastructures, Politecnico di Torino,C.so Duca degli Abruzzi, 24, 10129 Turin, Italye-mail: [email protected]
seasonal variability (Chi et al. 2009; Hiester et al. 1997;Moon et al. 2005; Ogura et al. 2001b; Rossini et al.2005a, b; Wallenhorst et al. 1997), dry and wet deposi-tion contributions (Pereira et al. 2007; Ren et al. 2007;Wang et al. 2010), particle deposition fluxes (Correaet al. 2006; Ogura et al. 2001a; Shih et al. 2006), typicaldeposition fluxes in urban, rural, and industrial areas(Correa et al. 2006; Ogura et al. 2001a; Pereira et al.2007; Rossini et al. 2005a, b; Van Lieshout et al. 2001;Wallenhorst et al. 1997; Wang et al. 2010), and theinfluence of single congeners and homologues (Chiet al. 2009; Correa et al. 2006; Moon et al. 2005; Renet al. 2007).
Our research aims at investigating the contribu-tion of a steel plant in northwestern Italy (700,000 tonsof steel/year) to the deposition fluxes of PCDD/Fs (interms of 17 most toxic 2,3,7,8-substituted PCDD/Fcongeners) at a local level, through the analysis ofmeasured data, then data in literature, and, finally,modeled data.
Few studies concerning predicted and measuredvalues of particle deposition (Holmes and Morawska2006; Huertas et al. 2012) are available. These studies,carried out both on coarse particles (Huertas et al. 2012;Kakosimos et al. 2011) and on fine particles (Huertaset al. 2012), highlight a significant correlation betweenmodeled and observed data.
The role of the same source to the local air concen-trations of PCDD/Fs was previously reported (Onofrioet al. 2011).
For the present study, a 15-month survey was con-ducted in 2005–2006 by the local EnvironmentalProtection Agency (ARPA Piemonte) in two monitoringstations that they defined to determine the depositionlevels in the vicinity of the steel plant, and simultaneous-ly in an unexposed sampling site, to evaluate the back-ground deposition fluxes.
The deposition fluxes measured in the surveys wereused to investigate the role of the source to the localdeposition fluxes, through the following:
– A comparison with emission data collected at thesource itself and data in literature;
– An analysis of congeners and homologues;– The principal component analysis (PCA).
The measured deposition fluxes were also comparedwith the pattern of PCDD/Fs in ambient air sampled atthe same sites (Fig. 1): the samples already reported in
our previous study were used for this analysis (Onofrioet al 2011).
The study was completed by identifying, through anair dispersion model (AERMOD), those areas expectedto have the highest deposition fluxes due to the steelplant (within a 420 km2 area surrounding the plant itself)and the order of magnitude of PCDD/F depositionfluxes referring to the steel plant.
Other studies highlighted the role and usefulness ofair dispersion model in evaluation of the contribution ofsources in local air pollution (Karademir 2004; Lee et al.2007; Rada et al. 2011).
Methods
Source characteristics
The source considered in this study was the air stackemissions from the flue gas treatment of a smelting steelplant (electric arc furnace—EAF) located about 40 kmfrom the city of Turin, in northwestern Italy. Detailedinformation about the source was previously reported(Onofrio et al. 2011). The main characteristics of thissource are the following:
– Stack inside diameter, 6.6 m;– Release height, 45 m;– Base elevation, 410 m s.l.m.;– UTM coordinate, (358800 EST; 4999100 NORD);– Gas exit flow rate, 1,100,000 Nm3 h−1;– Gas exit temperature, 45 °C;– Emission rate of PCDD/Fs (Ce), 0.05 ng I-TEQ
Nm−3;– Emission rate of total suspended particulate—
ed carbon (PAC) injected into a bag filter (filtrationvelocity 0.02 m/s).
The PCDD/Fs emission rate (Ce) and TSP emis-sion rate (Ct) were obtained through monitoring atthe chimney, which was carried out by the steel plantmanagement.
The mass concentration of TSP emitted by the sourcewas determined according to ISO 12141:2002 standard,using a manual gravimetric method.
As already specified in Onofrio et al (2011), both forsampling and analysis phases, UNI EN 1948 standard
Environ Monit Assess
was adopted to determine the PCDD/Fs concentration inthe flue gas and in TSP and the 2,3,7,8-substitutedPCDD/F congeners’ contribution both in vapor andsolid phases (UNI 1999a, b, c).
Table 1 shows the congener profiles and the averagepercentage distribution of PCDD/Fs obtained from thesamples collected at the source. The PCDD/Fs concen-tration is expressed in ng I-TEQ Nm−3 (NATO/CCMS1988), while the congener concentrations are expressedin ng Nm−3.
The samples collected showed a partitioning ofPCDD/Fs between vapor/solid phases of 95.5 % invapor phase and 4.5 % in solid phase.
Deposition sampling
In order to evaluate the PCDD/F deposition fluxes in thearea surrounding the steel plant, local EnvironmentalProtection Agency (ARPA Piemonte) set up two moni-toring stations: 1B (1.71 km northwest of the plant) and2VI (2.84 km southeast of the plant). The monitoringpoints were set up according to valley orientation and tothe prevailing wind direction (Fig. 3). Both the
monitoring points are located in rural areas. However,these two monitoring points and the source are influ-enced by the presence of an international highway andrail infrastructures (motorway and railway) and by thepresence of small metallurgical factories (the source isthe only steel plant in the area). An additional station(6GR) was set up 31.27 km from the steel plant, in orderto estimate the deposition fluxes outside the area ofinfluence of the source.
The position of the background sampling point 6GRwas established by ARPA Piemonte and located in anarea outside the influence of the source (a preliminaryair dispersion simulation identified this area) and ingeneral outside the influence of large industrial plants.At the same time, 6GR had the same characteristic bothenvironmental and anthropic of the area of influence ofthe source (mainly vehicular traffic and domesticheating).
Stations 1B, 2VI, and 6GR concur with those report-ed in Onofrio et al (2011).
The samplers were polymer structures formed by acylindrical container and a protection ring to avoiddamage by birds and animals. Each bulk had a
Fig. 1 PCDD/Fs deposition and air concentration homologue profiles
Environ Monit Assess
deposition surface of 0.038 m2. The structure wasclamped to a 60-mm pole. A Pyrex bottle with a funnelproperly silanized was placed in the support.
The samplers used were able to collect both dry andwet depositions (ARPA Piemonte 2007).
The limitation of this kind of sampling device is thatit is not suitable for dry gases which, according toSchröder et al. (1997), are of minor importance forPCDD/Fs (Lohmann and Jones 1998).
The samples were analyzed by ARPA Piemonte ac-cording to the Environmental Protection Agency (EPA)method 1613/94 (EPA 1994) using isotope dilution,high-resolution gas chromatography (HRGC) for theseparation, coupled with high-resolution mass spec-trometry (HRMS) for the quantification of 2,3,7,8-substituted PCDD/F congeners.
Nine samples were collected at three different sam-pling sites; Table 2 presents the monitoring periods andthe number of samples for each monitoring point. Foreach monitoring point, ARPA Piemonte establishedboth the sampling period and the type of deposition(total or dry) to be measured.
In each monitoring station, different numbers ofsamples were collected between October 2005 and
December 2006: four samples in 1B, three samplesin 2VI, and two samples in 6GR. The source emit-ted at a constant rate throughout the sampling pe-riod. The samples were collected both in winter andsummer, in order to identify seasonal variations inPCDD/Fs depositions (Chi et al. 2009; Hiester et al.1997; Moon et al. 2005; Ogura et al. 2001b;Rossini et al. 2005a, b; Wallernhost et al. 1997).The deposition samplings were also studied throughthe PCA.
The PCA was used to compare differing PCDD/Fspatterns among monitoring stations and source (11observations, 17 variables). Prior to modeling, thedataset was normalized using a centered logratiotransformation (Aitchison 2003) and then the datawas processed using the XLSTAT statistical analysissoftware.
Deposition fluxes modeling
The PCDD/F deposition fluxes modeling was imple-mented with AERMOD, a steady-state, plume disper-sion model developed by the US EPA. Previous studiesconfirmed the usefulness of AERMOD in providing
TOTAL (ng I-TEQ Nm−3) 5.47E-02 7.12E-02 6.30E-02 –
Environ Monit Assess
Table2
PCDD/Fsmeasureddepositio
nfluxes
ineach
monito
ring
station
Monito
ring
station
1B2V
I6G
R
Distancefrom
source,1.71km
;Num
berof
samples,4
Distancefrom
source,2.84km
;Num
berof
samples,3
Distancefrom
source,31.27
km;
Num
berof
samples,2
Samplingperiod
11-O
ct-05to
2-Feb-06
14-apr-06to
7-Jul-06
11-Jul-06to
18Sep-06
21-N
ov-06to
20-D
ec-06
2-Feb-06
to14-A
pr-06
14-apr-06to
7-Jul-06
21-N
ov-06to
20-D
ec-06
18-N
ov-05to
1-Feb-06
22-M
ay-06to
28-A
ug-06
Period
(day)
114
8469
2971
8429
7598
Precipitatio
n(m
m)
269.4
146.6
215.2
36.4
59.6
147.2
36.4
252.4
102.8
Average
Temperature
(°C)
2.3
17.1
21.1
4.5
5.2
17.2
4.5
−1.2
20.9
Congener
Totalfluxes
(pgm
−2day−
1)
Totalfluxes
(pgm
−2day−
1)
Dry
fluxes
(pgm
−2day−
1 )Dry
fluxes
(pgm
−2day−
1 )To
talfluxes
(pgm
−2day−
1)
Totalfluxes
(pgm
−2day−
1)
Dry
fluxes
(pgm
−2day−
1)
Totalfluxes
(pgm
−2day−
1)
Dry
fluxes
(pgm
−2day−
1)
2.3.7.8-TCDD
0.005
0.005
0.01
0.015
0.005
0.005
0.015
0.01
0
1.2.3.7.8-PeCDD
0.015
0.01
0.015
0.015
0.015
0.01
0.015
0.015
0.75
1.2.3.4.7.8-HxC
DD
0.005
0.015
0.01
0.015
0.01
0.005
0.015
0.01
0.005
1.2.3.7.8.9-HxC
DD
0.01
1.13
0.01
0.015
0.01
0.005
0.02
0.015
0.64
1.2.3.6.7.8-HxC
DD
0.37
0.015
0.01
0.015
0.01
0.005
0.02
0.015
0.005
1.2.3.4.6.7.8-HpC
DD
2.45
21.16
7.32
0.01
3.48
411.8
3.65
2.79
OCDD
6.19
116.23
40.89
21.06
9.27
19.63
27.95
15.09
13.59
2.3.7.8-TCDF
1.15
1.69
2.67
3.18
1.48
0.015
3.9
1.68
0.005
2.3.4.7.8-PeCDF
1.52
0.01
3.2
3.72
2.15
1.81
3.18
1.26
0.005
1.2.3.7.8-PeCDF
0.88
2.07
1.465
1.72
1.41
1.56
0.01
1.05
1.02
1.2.3.4.7.8-HxC
DF
1.02
0.01
2.06
2.36
1.71
1.06
2.27
0.77
0.005
1.2.3.7.8.9-HxC
DF
0.01
0.015
0.015
0.01
0.015
0.01
0.015
0.005
0.005
1.2.3.6.7.8-HxC
DF
0.69
1.63
0.01
2.63
1.33
0.01
2.27
0.7
0.005
2.3.4.6.7.8-HxC
DF
0.83
0.01
0.015
2.81
1.93
1.19
31.05
1.29
1.2.3.4.6.7.8-HpC
DF
1.66
0.005
5.49
6.72
3.63
310.26
2.04
2.52
1.2.3.4.7.8.9-HpC
DF
0.28
0.8
77.97
0.005
0.01
0.01
1.18
0.005
0.005
OCDF
0.97
5.07
5.04
5.81
2.45
2.13
9.71
1.4
3.49
TOTA
L(pgm
−2day−
1 )18.06
149.88
146.20
50.11
28.92
34.46
75.63
28.77
26.13
TOTA
L(pgI-TEQm
−2day−
1)
1.28
0.91
3.12
3.17
1.89
1.31
3.03
1.20
0.69
Environ Monit Assess
information both for ambient air and deposition expo-sure assessment purposes (Barton et al. 2010;Krzyzanowski 2011).
In this paper, the AERMOD model was initializedwith the ISC-AERMOD View 6.7.1 version.
The meteorological hourly data used for the sim-ulations was provided by the local EnvironmentalProtection Agency (ARPA Piemonte) and measuredin the Borgone di Susa monitoring station ( ARPAPiemonte 2005; 2006), 3.6 km southeast from thesteel plant and 400 m above sea level.
The local orographic conditions were obtainedthrough the automated download of SRTM3 terraindata files (90 m resolution) in AERMAP terrainprocessor.
According to the samples collected by ARPAPiemonte, trough AERMOD were calculated:
& Dry deposition flux of PCDD/Fs: determined as thesum of dry particulate deposition and dry gasdeposition;
& Total deposition flux of PCDD/Fs : determined asthe sum of dry and wet particulate deposition anddry and wet gas deposition.
Regarding particle deposition, on the basis ofthe particle size distribution of the samples collect-ed at the source, a fine particle fraction of 90 %and a mass mean particle diameter of 0.5 μm wasset up.
This setting was consistent with previous studies thatoutlined that at least 75 % of the PCDD/Fs are attachedto the particles whereas the remaining part can be foundin the gas phase (Kurokawa et al. 1996, 1998; Shih et al.2009).
Furthermore, other studies outlined the difficulty incharacterizing the general distribution of PCDD/Fs withparticle size (Oh et al. 2002); however, the largestamount of PCDD/Fs is concentrated on fine particles(<1 μm) (Kurokawa et al. 1998).
In our study, PCDD/Fs particulate depositionwas determined by setting the model emission rateat Ct (0.035 mg TSP Nm−3) and selecting method2, where less than 10 % of TSP has a diameterlarger than 10 μm. Furthermore, the PCDD/Fsparticulate deposition was determined by applyingthe PCDD/Fs concentration on the TSP measured at thesource to the modeled TSP deposition flux calculated byAERMOD.
On the other hand, the PCDD/Fs gas deposition wasdetermined as follows:
& Setting the emission rate at Ce (0.05 ng PCDD/Fs I-TEQ Nm−3);
& Setting the physical parameters (diffusivity in air,diffusivity in water, cuticular resistance, andHenry’s Law constant) for each dioxin and furanfound in EPA’ s database;
& Using default gas deposition parameters suggestedby AERMOD;
& Setting “Urban Land” as land-use category. Thisland-use category was selected in order to comparethe modeling results with the bulk depositionsamplings.
In order to model the total deposition, AERMODcalculated both PCDD/Fs particle deposition and gasdeposition in dry and wet conditions. The parametersfor calculation in both the conditions were available inhourly meteorological data processed by AERMET.
The simulations were conducted both on discrete anduniform receptors.
The discrete receptors had the same coordinates asthe monitoring stations (1B, 2VI, 6GR).
The uniform Cartesian receptors were set up in a 420-km2 area around the source, converted into a 100 m×50 m raster grid (a total of 84,600 receptors).
Results and discussion
Deposition sampling
ARPA Piemonte assessed the PCDD/F depositionfluxes through three bulk samplers located in thethree different monitoring stations (1B, 2VI, and6GR) in the period between October 2005 andDecember 2006.
Each monitoring station had a different monitoringplan (different monitoring months and number of sam-plings) that is shown in Table 2.
Table 2 also shows the results of the monitoringcarried out by ARPA Piemonte and the analytic partitionand distribution of PCDD/Fs in 17 congeners. PCDD/Fdeposition fluxes were reported both in picograms persquare meter per day and in picograms I-TEQ per squaremeter per day, while the congener contributions areexpressed in picograms per square meter per day Table 2.
Environ Monit Assess
The PCDD/F-measured deposition fluxes had the sameorder of magnitude in all the monitoring stations. In 1Band 2VI, the measured deposition fluxes were respectivelybetween 0.91 pg I-TEQm−2 day−1 and 3.17 pg I-TEQm−2
day−1 in 1B, between 1.31 pg I-TEQ m−2 day−1 and3.03 pg I-TEQm−2 day−1 in 2VI. Lower deposition valuesweremeasured in the unexposed sampling site 6GR (0.69–1.20 pg I-TEQm−2 day−1), outside the sphere of influenceof the source.
In all the monitoring stations, Octachloro dibenzo-p-dioxin (OCDD) dominated the PCDD/Fs depositionfluxes. Among PCDDs, heptachlorodibenzo-p-dioxin(HpCDD) also showed significant levels, while otherhomologues were almost absent. Among PCDFs, thepredominant homologues were octachlorodibenzofuran(OCDF), scrivere heptachlorodibenzofuran (HpCDFs),and Hexachlorodibenzofuran (HxCDFs), but less chlo-rinated homologues also had levels which were notnegligible. These results are consistent with other stud-ies (Chi et al. 2009; Correa et al. 2006; Guerzoni et al.2004; Ogura et al. 2001b).
In terms of single congeners, the measured depo-sition fluxes showed that the most prominent indi-vidual congeners were OCDD, 1,2,3,4,6,7,8-HpCDD,1,2,3,4,6,7,8-HpCDF, and OCDF. These congenerprofiles are similar to those of other findings(Koester and Hites; 1992; Lohmann and Jones1998; Ren et al. 2007; Shih et al. 2006; Viveset al. 2008; Wang et al. 2010).
This evidence suggests the presence of further sourcesoutside the steel plant, as it has low levels of 1,2,3,4,6,7,8HpCDD, OCDD, 1,2,3,4,6,7,8 HpCDF, and OCDF andit is characterized by tetrachlorodibenzofuran (TCDF)and pentachlorodibenzofurans (PeCDFs).
In relation to seasonality, according to other studies(Chi et al. 2009; Lohmann and Jones 1998; Moon et al.2005; Ogura et al. 2001a; Rossini et al. 2005a, b) thePCDD/F deposition fluxes measured in winter samples(October to March) were higher than those measured insummer samples (April to September) in all the moni-toring stations (1B, 2VI, and 6GR).
Furthermore, the deposition fluxes of PCDD/Fsincreased as the temperature decreased and thedifferences in fluxes depending on the ambient temper-ature were larger for less chlorinated PCDD/Fs com-pared to more chlorinated PCDD/Fs (Moon et al. 2005;Ogura et al. 2001a, b). In particular, TCDF showed ahigher contribution to the total deposition in wintersamples (6.4 % in 1B, 5.1 % in 2VI, and 5.8 % in
6GR) compared to summer ones (1.5 % in 1B, 0.05 %in 2VI, and 0.02 % in 6GR).
Similar results were also highlighted for PeCDFs in1B and 6GR (respectively, 7.92 and 4.38 % in winterand 1.1 and 0.02 % in summer), while TCDD andPeCDD sampled values were so low that no consider-ations can be made. No correlation was found betweenthe deposition fluxes and the amount of precipitation(Ogura et al. 2001a).
Within the study, a comparison between the patternof PCDD/Fs of the air concentration sampling (Onofrioet al. 2011) and the deposition sampling, collected con-currently at the same sites, was conducted. The compar-ison showed a similarity between the air concentrationand deposition pattern of PCDD/Fs, only OCDD in-creased its contribution significantly (from 23 % in theair concentration to 45 % in the deposition). This resultis consistent with those given by in other studies(Lohmann and Jones 1998).
Figure 1 shows both the deposition and air concen-tration homologue profiles in the three sampling sitesand at the source. They are expressed in terms of relativecontribution of single homologue. The graph includeserror bars representing single standard deviations thatindicate a significant variability in the relative abun-dance of the different homologues.
The PCDD/F deposition fluxes measured in the threemonitoring stations were consistent with other studiescarried out in Italy in urban and suburban areas(Guerzoni et al. 2004; Rossini et al. 2005a, b) and in ruralEuropean areas (Wallenhorst et al. 1997); while they arelower than those measured in other European urban/ sub-urban areas (Halsall et al. 1997; Van Lieshout et al. 2001;Wallenhorst et al. 1997) or in sites influenced by industrialsources (Ren et al. 2007; Rossini et al. 2005a, b).
Monitoring campaigns carried out in Asian countriesshowed PCDD/Fs deposition fluxes sensibly higherthan those monitored in Italy and in other Europeancountries (Chi et al. 2009; Moon et al. 2005; Oguraet al. 2001a, b; Ren et al. 2007; Wang et al. 2010).
Table 3 lists the literature data concerning PCDD/Fdeposition fluxes outlined in other studies carried out inseveral international areas.
Figure 2 shows the resulting score plot of the PCA,with the samples included in the analysis shown in thespace of the first and second component. The firstcomponent takes into account 35 % of the variabilityof the dataset and was mainly influenced by lowerchlorinated homologues (tetra-, penta-, hexa-CDD, and
Environ Monit Assess
Table3
Literature
PCDD/Fsdepositio
nfluxes
Reference
Location
Urban–suburbanareas
Industrialareas
Ruralareas
Halsalletal.(1997)
UK
Urban
Cardiff,375–2060(ngm
−2year−1);
Urban
Manchester,360–2,200(ngm
−2year−1)
Wallenhorstetal.(1997)
Germany
Urban,12.4–14.9(ngTEQm
−2year−1);
Suburban,8.8–13.9(ngTEQm
−2year−1)
1.0–3.7(ng
TEQm
−2year−1)
Van
Lieshoutetal.(2001)
Belgium
Urban,1.24–9.12
(ngTEQm
−2year−1)
0.25–4.0(ngTEQm
−2year−1)
Ogura
etal.(2001a)
Japan
Urban
Tokyo,800–2,200(ngm
−2year−1);
Urban
Yokoham
a,440–1,800(ngm
−2year−1);
Suburban
Tsukuba,410–4,300(ngm
−2year−1)
160–980(ngm
−2year−1)
Guerzonietal.(2004)
Italy
Urban:∼
50(pgm
−2day−
1);Su
burban:
∼50(pgm
−2day−
1)
∼100
(pgm
−2day−
1)
Moonetal.(2005)
Korea
Urban,68–228(ngm
−2year−1);
Suburban,38–252(ngm
−2year−1)
Rossini
etal.(2005a,b)
Italy
Urban
Mestre,46–169
(pgm
−2day−
1);
Urban
Venice,13–200
(pgm
−2day−
1)
Porto
Marghera,67–2,767
(pgm
−2day−
1);About
5km
ofPo
rtoMarghera,15–431
(pgm
−2day−
1)
Correaetal.(2006)
Texas(U
S)527(pgm
−2day−
1)
Pereiraetal.(2007)
Brazil
0.10–1.9
(pgWHO-TEQm
−2day−
1)
winter;0.96–2.2(pgWHO-TEQ
m−2
day−
1)summer
0.10–1.2(pgWHO-TEQ
m−2
day−
1)winter;0.11–2.1
(pgWHO-TEQm
−2day−
1)
summer
Ren
etal.(2007)
China
Urban
Wushan,59–1,640
(pgm
−2day−
1);
Urban
Haizhu,440–770(pgm
−2day−
1);
Suburban
Changban,380–1,090(pgm
−2day−
1)
536–1,220(pgm
−2day−
1)
Chi
etal.(2009)
NorthernTaiwan
106–219(pgm
−2day−
1)winter;95.1–155
(pgm
−2day−
1)summer
Wangetal.(2010)
Taiwan
Suburban,168(ngm
−2year−1)
310(ngm
−2year−1)
Coastalarea,135
(ngm
−2year−1);Agricultural
area,115
(ngm
−2year−1)
Environ Monit Assess
tetra-, penta-CDF) in the positive direction and byhigher chlorinated homologues (hepta- and octa-CDDand hepta- and octa-CDF) in the negative direction. Thesecond component accounts for 24 % of the variabilityof the data set and was mainly influenced by tetra-CDD, tetra-, and hexa-CDF in positive direction;penta-, hexa- , hepta- and octa- CDD, and octa-CDF innegative direction.
The PCA analysis made a clear distinction betweenthe profiles of the monitoring stations and the profiles ofthe source. The samples collected at the source hadconcentration profiles in the positive direction of thefirst and second component, while the profiles of themonitoring stations mainly fell within a cluster in thenegative direction of the first component and in thepositive direction of the second one.
AERMOD modeling
In order to have further elements to evaluate the influ-ence of the steel plant on the local deposition fluxes ofPCDD/Fs and to identify the areas characterized by thehighest PCDD/Fs deposition fluxes, these fluxes were
calculated by AERMOD in the area surrounding theplant.
In particular, AERMOD simulated both gas and par-ticle deposition fluxes, on two type of receptors:
– Uniform Cartesian grid: AERMOD calculated thePCDD/F deposition fluxes for each of the 84,600receptors set up in an area also containing the twodiscrete receptors (1B, 2VI) and the unexposedsampling site (6GR). The simulation was carriedout using 1 year of meteorological data, fromJanuary 1st 2006 to December 31st 2006. Figure 3shows the total deposition map (interpolated modeloutput of deposition fluxes) obtained for the aboveaverage period;
– Discrete receptors: in 1B, 2VI, and 6GR the simu-lations were carried out using the same monitoringperiod shown in Table 2.
An AERMOD 1-year simulation outlined a maxi-mum PCDD/Fs deposition area located within 2 kmfrom the source and in west–northwestern direction,both according to the prevalent wind direction and to
1B1
1B2
1B3
1B4
2VI1
2VI2
2VI3
6GR1S1
S2
6GR2
-5
-4
-3
-2
-1
0
1
2
3
Seco
nd c
ompo
nent
(24
%)
First component (35 %)
PCA BULK DEPOSITION
-3 -2 -1 0 1 2 3 4 5 6
Fig. 2 PCA score plot
Environ Monit Assess
the orientation of the valley where the source is located.Inside this area, the maximum value of the PCDD/Ftotal deposition calculated by AERMOD was 282 pgI-TEQ m−2 year−1, one order of magnitude higher than
in 1B and 2VI, and two orders of magnitude higher thanin 6GR.
The monitoring station closest to the maximum de-position area was 1B. In 1B, AERMOD calculated a
PICOGRAMS/M**2
1B
2VI
6GR
N
Monitoring station 1B 2VI 6GR Max Value 6002 )raey( doireP
a Deposition fluxes too low to be calculated by AERMOD
Environ Monit Assess
deposition flux related to the source of 33.67 pg I-TEQm−2 year−1.
In the monitoring station 2VI, that is located outsidethe maximum deposition area, AERMOD calculated adeposition flux related to the source of 14.69 pg I-TEQm−2 year−1, while in 6GR the deposition flux related tothe source was reduced to 1.39 pg I-TEQ m−2 year−1.
The modeling was repeated using the monitoring pe-riod of each monitoring station, listed in the second rowof Table 2, in order to compare the measured depositionvalues to AERMOD output values for each receptor (1B,2VI, and 6GR). Both these values are shown in Table 4.
In both monitoring points (1B, 2VI), AERMOD cal-culated deposition fluxes of two or three orders of mag-nitude lower than those measured at the same points.
The AERMOD simulations confirmed that point6GR was not subject to emissions from the steel plant,in fact the model was not able to calculate the depositionfluxes at this point because they were very low andunder the limit of sensitivity of the model itself.
AERMOD confirmed the low influence of the sourceinvestigated in the PCDD/Fs local deposition fluxes andthe results given by the analysis of the measured depo-sition fluxes, the PCA analysis, and the study of thehomologue profiles of the samples collected.
Conclusion
The aim of this paper was to identify the contribution ofa steel plant located in northwestern Italy to thePCDD/Fs deposition fluxes at a local level.
The study was carried out using deposition samplesmeasured in three points around the steel plant (twomonitoring stations and one background station), datain literature, and AERMOD simulation results.
In all the monitoring stations (1B, 2VI, and 6GR), themeasured deposition fluxes of the PCDD/Fs had thesame order of magnitude, with lower values measuredin the background station 6GR (0.91–3.17 pg I-TEQm−2 day−1 in 1B; 1.31–3.03 pg I-TEQm−2 day−1 in 2VI,and 0.69–1.20 pg I-TEQ m−2 day−1 in 6GR).
The comparison between the PCDD/Fs depositionfluxes measured and data in literature showed that thetotal measured deposition fluxes of PCDD/Fs (in termsof 17most toxic 2,3,7,8-substituted PCDD/F congeners)were consistent with studies carried out in Italy in urbanand suburban areas and in rural European areas; whilethese were lower than those measured in other European
urban/suburban areas or in sites influenced by industrialsources.
The comparison between homologue profile of thesource and of deposition and air concentration samplesshowed a similarity between air concentration and depo-sition patterns of PCDD/Fs and a clear distinction betweenthe source and air concentration and deposition patterns.
These results were also confirmed by PCA analysis.An AERMOD 1-year simulation highlighted a max-
imum PCDD/Fs deposition area located within 2 kmfrom the source and in west–northwestern direction andall the monitoring stations fell outside this area. Thesimulation was also conducted for discrete receptors: in1B and 2VI, AERMOD calculated deposition fluxes oftwo or three orders of magnitude lower than those mea-sured at the same points and confirmed also that point6GR was not subject to emissions from the steel plant.
The analysis of measured data, then data in literature,and, finally, modeled data highlighted the limited influ-ence of the source in the local PCDD/F depositionfluxes and confirmed the results of a previous studyconcerning the role of the same source in the localPCDD/F air concentrations.
The measured data also confirmed that TCDF couldbe used as a tracer of the pollutants emitted by the source(steel plant). In fact, it was present at the source in highpercentages, while it was almost absent at all the mon-itoring points, both in air concentration and depositionfluxes.
Furthermore, AERMOD highlighted different maxi-mum fallout areas respectively for air concentrationsand deposition fluxes, in terms of number, size, anddistance from the source. This aspect confirmed theusefulness of AERMOD as a complementary tool topredict the maximum fallout areas and to plan the cor-rect placement of a monitoring network, in order toinvestigate the contribution of a source to the localpollution. A wrong placement of a monitoring networkcould in fact lead to the underestimation of the exposurelevels of the local dwellers.
References
Aitchison, J. (2003). The statistical analysis of compositional data(2nd ed.). Caldwell, New Jersey: The Blackburn Press.
ARPA Piemonte .(2005-2006). Banca dati meteorologica.<http://www.arpa.piemonte.it/index.php?module=ContentExpress&func=display&btitle=CE&mid=&ceid=389>.
Piemonte, A. R. P. A. (2007). Analisi di rischio connesso con lostato di contaminazione ambientale nel territorio della BassaValle di Susa. Turin, Italy: ARPA Piemonte (in Italian).
Bakoglu, M., Karademir, A., & Durmusoglu, E. (2005).Evaluation of PCDD/F levels in ambient air and soils andestimation of deposition rates in Kocaeli, Turkey.Chemosphere, 59, 1373–1385.
Barton, C. A., Zarzecki, C. J., & Russell, M. H. (2010). A site-specific screening comparison of modeled and monitored airdispersion and deposition for perfluorooctanoate. Journal ofthe Air & Waste Management Association, 60, 402–411.
Benfenati, E., Valzacchi, S., Mariani, G., Airoldi, L., & Fanelli, R.(1992). PCDD, PCDF, PCB, PAH, CADMIUM and LEADin roadside soil: relationship between road distance and con-centration. Chemosphere, 24, 1077–1083.
Chetwittayachan, T., Shimazaki, D., & Yamamoto, K. (2002). Acomparison of temporal variation of particle-bound polycy-clic aromatic hydrocarbons (pPAHs) concentration in differ-ent urban environments: Tokyo, Japan, and Bangkok,Thailand. Atmosphere Environment, 36, 2027–2037.
Chi, K. H., Liu, K. T., Chang, S. H., & Chang, M. B. (2009).Atmospheric deposition of PCDD/Fs measured via automatedand traditional samplers in Northern Taiwan. Chemosphere,77, 1184–1190.
Colombo, A., Benfenati, E., Mariani, G., Lodi, M., Marras, R.,Rotella, G., et al. (2009). PCDD/Fs in ambient air in north-east Italy: the role of a MSWI inside an industrial area.Chemosphere, 77, 1224–1229.
Correa, O., Raun, L., Rifai, H., Suarez, M., Holsen, T., & Koenig,L. (2006). Depositional flux of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans in an urban set-ting. Chemosphere, 64, 1550–1561.
EPA. (1994). Tetra- through octa-chlorinated dioxins and furansby isotope dilution HRGC/HRMS. Method 1613/94.
Guerzoni, S., Rossini, P., Molinaroli, E., Rampazzo, G., &Raccanelli, S. (2004). Measurement of atmosphericdeposition of polychlorinated dibenzo-p-dioxins and diben-zofurans in the Lagoon of Venice, Italy. Chemosphere, 54,1309–1317.
Halsall, C. J., Coleman, P. J., & Jones, K. C. (1997). Atmosphericdeposition of polychlorinated dibenzo-p-dioxins/dibenzofu-rans (PCDD/Fs) and polycyclic aromatic hydrocarbons(PAHs) in two UK cities. Chemosphere, 35, 1919–1931.
Hiester, E., Bruckmann, P., Bohm, R., Eynck, P., Gerlach, A.,Mulder, W., et al. (1997). Pronounced decrease ofPCDD/PCDF burden in ambient air. Chemosphere, 34,1231–1243.
ISO. (2002). Stationary source emissions—determination of massconcentration of particulate matter (dust) at low concentra-tions—manual gravimetric method. ISO 12141:2002
Holmes, N. S., & Morawska, L. (2006). A review of dispersionmodeling and its application to the dispersion of particles: anoverview of different dispersion models available.Atmospheric Environment, 40, 5902–5928.
Huertas, J. I., Huertas, M. E., & Izquierdo, S. (2012). Air qualityimpact assessment ofmultiple open pit coal mines in northernColombia. Journal of Environmental Management, 93, 121–129.
Kakosimos, K. E., Assael, M. J., & Katsarou, A. S. (2011).Application and evaluation of AERMOD on the assessmentof particulate matter pollution caused by industrial activities
in the Greater Thessaloniki area. Environmental Technology,32, 593–608.
Karademir, A. (2004). Health risk assessment of PCDD/Femissions from a hazardous and medical waste incin-erator in Turkey. Environmental International, 30, 1027–1038.
Koester, C. J., & Hites, R. A. (1992). Wet and dry deposition ofchlorinated dioxins and furans. Environmental Science andTechnology, 26, 1375–1382.
Krzyzanowski, J. (2011). Approaching cumulative effects throughair pollution modelling. Water Air Soil Pollution, 214, 253–273.
Kurokawa, Y., Matsueda, T., Nakamura, M., Takada, S., &Fukamachi, K. (1996). Characterization of non-ortho copla-nar PCBs, polychlorinated dibenzo-p-dioxins and dibenzofu-rans in the atmosphere. Chemosphere, 32, 491–500.
Kurokawa, Y., Takahiko, M., Matayoshi, N., Satoshi, T., &Kazumi, F. (1998). Distribution of polychlorinated dibenzo-p-dioxins and dibenzofurans in various sizes of airborneparticles. Chemosphere, 37, 2161–2171.
Lee, S. J., Choi, S. D., Jin, G. Z., Oh, J. E., Chang, Y. S., & Shin, S.K. (2007). Assessment of PCDD/F risk after implementationof emission reduction at a MSWI. Chemosphere, 68, 856–863.
Li, Y., Wang, P., Ding, L., Xiaomin, L., Wang, T., Zhang, Q., et al.(2010). Atmospheric distribution of polychlorinated dibenzo-p-dioxins, dibenzofurans and dioxin-like polychlorinated bi-phenyls around a steel plant area, Northeast China.Chemosphere, 79, 253–258.
Lohmann, K., & Jones, K. C. (1998). Dioxins and furans in air anddeposition: a review of levels, behaviour and processes. TheScience of the Total Environment, 219, 53–81.
Mari, M., Nadal, M., Schuhmacher, M., & Domingo, J. L. (2008).Monitoring PCDD/Fs, PCBs and metals in the ambient air ofindustrial area of Catalonia, Spain. Chemosphere, 73, 990–998.
Moon, H. B., Lee, S. J., Choi, H. G., & Ok, G. (2005).Atmospheric deposition of polychlorinated dibenzo-p-di-oxins (PCDDs) and dibenzofurans (PCDFs) in urban andsuburban areas of Korea. Chemosphere, 58, 1525–1534.
NATO/CCMS. (1988). International toxicity equivalency factor (I-TEF) method of risk assessment for complex mixtures ofdioxins and related compounds. Pilot Study on InternationalInformation Exchange on Dioxins and Related Compounds,Report Number 176, August 1988, North Atlantic TreatyOrganization. Committee on the Challenges of ModernSociety.
Ogura, I., Masunaga, S., & Nakanishi, J. (2001a). Atmosphericdeposition of polychlorinated dibenzo-p-dioxins,polychlorinated dibenzofurans, and dioxin-like polychlorinatedbiphenyls in the Kanto Region, Japan. Chemosphere, 44,1473–1487.
Ogura, I., Masunaga, S., & Nakanishi, J. (2001b). Congener-specific characterization of PCDDs/PCDFs in atmosphericdeposition: comparison of profiles among deposition,source, and environmental sink. Chemosphere, 45,173–183.
Oh, J. E., Chang, Y. S., Kim, E. J., & Lee, D. W. (2002).Distribution of polychlorinated dibenzo-p-dioxins and diben-zofurans (PCDD/Fs) in different sizes of airborne particles.Atmospheric Environment, 36, 5109–5117.
Environ Monit Assess
Onofrio, M., Spataro, R., & Botta, S. (2011). The role of a steelplant in north-west Italy to the local air concentrations ofPCDD/Fs. Chemosphere, 82, 708–717.
Pereira, M. D., Waller, U., Reifenhauser, W., Reifenhäuser,W., Torres, J. P. M., Malm, O., et al. (2007). Persistentorganic pollutants in atmospheric deposition and bio-monitoring with Tillandsia Usneoides (L.) in an indus-trialized area in Rio de Janeiro state, south eastBrazil—Part I: PCDD and PCDF. Chemosphere, 67, 1728–1735.
Rada, E. C., Ragazzi, M., Zardi, D., Laiti, L., & Ferrari, A. (2011).PCDD/F enviromental impact from municipal solid wastebio-drying plant. Chemosphere, 84, 289–295.
Ren, M., Peng, P., Zhang, S. K., Yu, L., Zhang, G., Mai, B., et al.(2007). Atmospheric deposition of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) in Guangzhou,China. Atmospheric Environment, 41, 592–605.
Rossini, P., Guerzoni, S., Matteucci, G., Gattolin, M., Ferrari, G.,& Raccanelli, S. (2005a). Atmospheric fall-out of POPs(PCDD-Fs, PCBs, HCB, PAHs) around the industrial districtof Porto Marghera, Italy. Science Total Environment, 349,190–200.
Rossini, P., Guerzoni, S., Molinaroli, E., Rampazzo, G., DeLazzari, A., & Zancanaro, A. (2005b). Atmospheric bulkdeposition to the lagoon of Venice: part I. Fluxes of metals,nutrients and organic contaminants. EnvironmentalInternational, 31, 959–974.
Schecter, A., Birnbaum, L., Ryan, J. J., & Constable, J. D. (2006).Dioxins: an overview. Environmental Research, 101, 419–428.
Schröder, J., WelschPausch, K., & McLachlan, M. S. (1997).Measurement of atmospheric deposition of polychlorinateddibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) to asoil. Atmospheric Environment, 31, 2983–2989.
Shih, M. L., Lee, W. S., Chang-Chien, G. P., Wang, L. C., Hung,C. Y., & Lin, K. C. (2006). Dry deposition of polychlorinateddibenzo-p-dioxins and dibenzofurans (PCDD/Fs) in ambientair. Chemosphere, 62, 411–416.
Shih, T. S., Shih, M., Lee, W. J., Huang, S. L., Wang, L. C., Chen,Y. C., et al. (2009). Particle size distributions and health-related exposures of polychlorinated dibenzo-p-dioxins anddibenzofurans (PCDD/Fs) of sinter plant workers.Chemosphere, 74, 1463–1470.
UNI. (1999a). Stationary source emissions—determination of themass concentration of PCDDs/PCDFs – Sampling, vol. UNIEN 1948–1:1999.
UNI. (1999b). Stationary source emissions—determination of themass concentration of PCDDs/PCDFs—extraction andclean-up, vol. UNI EN 1948–2:1999.
UNI. (1999c). Stationary source emissions—determination of themass concentration of Cs—identification and quantification,vol. UNI EN 1948–3:1999.
Van Lieshout, L., Desmedt, M., Roekens, E., De Fre, R., VanCleuvenbergen, R., & Wevers, M. (2001). Deposition ofdioxins in Flanders (Belgium) and a proposition for guidevalues. Atmospheric Environment, 35, 83–90.
Vives, I., Muller, A., Umlauf, G., Christoph, E. H., Mariani, G.,Skejo, H., et al. (2008). Levels of PCDD/Fs and trace ele-ments in superficial soils of Pavia Province (Italy).Environmental International, 34, 994–1000.
Wallenhorst, T., Krauβ, P., &Hangenmaier, H. (1997). PCDD/F inambient air and deposition in Banden-Württemberg,Germany. Chemosphere, 34, 1369–1378.
Wang, Y. F., Hou, H. C., Li, H. W., Lin, L. F., Wang, L. C., Chang-Chien, G. P., et al. (2010). Dry and wet depositions ofpolychlorinated dibenzo–dioxins and dibenzofurans in theatmosphere in Taiwan. Aerosol Air Quality Researsch, 10,378–390.
