Dry and Wet Deposition of Polychlorinated Dibenzo-p-dioxins and Dibenzofurans on the Drinking Water Treatment Plant

This study investigated the concentrations and congener profiles of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in intake source water (source water) and tap drinking water (tap water) of drinking water treatment plants (DWTPs). In order to have a better understanding on the influence of atmospheric PCDD/F deposition on drinking water, PCDD/Fs in the ambient air of a DWTP (DWTP-LN) were measured and both dry and wet deposition on the water treatment facilities were assessed. The results of this study indicated that the mean PCDD/F concentration in tap water (0.0039 pg WHO-TEQ/L) was found to be approximately 55% of magnitude less than that in source water (0.0086 pg WHO-TEQ/L). In addition, the total deposition flux (dry + wet) of PCDD/Fs entering the DWTP-LN was 27.0 ng I-TEQ/m-year, and wet and dry deposition contributed approximately 12.0% and 88%, respectively. It reveals that dry deposition is more important than wet deposition of PCDD/Fs in the ambient air of DWTP-LN. Atmospheric deposition of PCDD/Fs will increase the level in source water of DWTP-LN up to 8.91 × 10 pg I-TEQ/L, which is approximately 92% of the PCDD/Fs in source water. If a removal efficiency of 87% is achieved by conventional treatment processes including coagulation, flocculation, sedimentation and rapid sand filtration, the water after treatment may increase 1.16 × 10 pg I-TEQ/L, which is approximately 43% of the concentration level in tap water. These results indicate that in the DWTP-LN, the influence of atmospheric deposition of PCDD/Fs on the drinking water is of great significance, and water treatment facility with a cover is suggested.


INTRODUCTION
Due to the potential adverse health effects from exposure to polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs), such as reproductive difficulties and increased risk of cancer (US EPA, 2003), their presence in the environment is of great concern.In addition, PCDD/Fs are persistent in the environment, remaining intact for long periods (Stockholm Convention, 2001).Furthermore, PCDD/Fs are typically hydrophobic and lipophilic, thus in aquatic systems and soils they partition strongly to solids rather than the aqueous phase (Jones and de Voogt, 1999).Because of these properties, they become widely distributed, accumulate in the fatty tissue of living organisms and are toxic to humans and wildlife (Stockholm Convention, 2001).
In Japan, the environmental quality standard for dioxins in water is 1 pg WHO-TEQ/L or less (Environmental Health Department, 2002).While in the USA, the highest level of 2,3,7,8-TCDD that is allowed in drinking water is 3 × 10 -8 mg/L (equal to 30 pg/L), which was set by US EPA (2003).PCDD/Fs enter the aquatic environment can occur with the discharge of wastewater (UNEP Chemicals, 2005), application of pesticides (Masunaga et al., 2001;UNEP Chemicals, 2005) as well as dry and wet deposition from the atmosphere.But how PCDD/Fs enter tap water through deposition has seldom been investigated, a deficiency this study aims to address.
In Taiwan, there were several studies regarding the atmospheric PCDD/F deposition from relevant sources.In the study of Chi et al. (2009), owing to the automated PCDD/F ambient sampler can prevent both re-suspension and photo degradation of the PCDD/Fs collected, the PCDD/F deposition flux collected using the automated PCDD/F sampler was significantly higher than that sampled with the cylindrical vessels.The wet deposition flux of PCDD/Fs observed was significantly higher than the dry deposition flux, demonstrating that wet deposition is the major PCDD/F removal mechanism in the atmosphere.Shih et al. (2006) investigated dry deposition of PCDD/Fs in the ambient air in southern Taiwan.Atmospheric dry deposition fluxes of total PCDD/Fs averaged approximately 150 pg/m 2 /day.The total dry deposition flux was found to decrease as the temperature increased.Calculated dry deposition velocities of total PCDD/Fs were averaged 0.42 cm/s.Atmospheric dry deposition of PCDD/Fs in the vicinity of municipal solid waste incinerators was investigated by Wu et al. (2009).Dry deposition fluxes of total PCDD/Fs were 18.0 and 23.5 pg I-TEQ/m 2 -day in the ambient air near two MSWIs.They were considerably higher than those measured in Guangzhou, China.While annual dry deposition fluxes were 189 and 217 ng/m 2 -year, which were also much higher than that near to the Atlantic Ocean.Notably, Asian dust storm (ADS) that originated in the deserts of Mongolia and Mainland China eventually reached Taiwan, and then significantly increased the atmospheric PCDD/F concentrations.Additionally, the amount of PCDD/Fs bound to suspended particles increased during the ADS episode (Chi et al., 2008).
In this study, the data for PCDD/Fs in intake source water, tap water, and the ambient air on a water treatment plant are investigated.In addition, atmospheric dry-and wet-deposition on the surface of water treatment facilities was assessed and its influence on the level of PCDD/Fs in treated water was evaluated.

PCDD/Fs in Source Water and Tap Water
In order to have a better understanding of PCDD/Fs in source water of DWTPs in Taiwan, a total of 17 samples were taken from a reservoir (Sin-Shan Reservoir, n = 1) located in northern Taiwan, a river located in middle Taiwan (Jhuo-Shuei River, n = 4), and a river located in southern Taiwan (Gao-Ping River, n = 12).Jhuo-Shuei River is the longest river in Taiwan, and Gao-Ping River has the largest watershed area in Taiwan.
For tap water, a total of 4 samples of treated water taken from a drinking water treatment plant (DWTP-LN) using the water of Jhuo-Shuei River as raw water.In addition, a total of 14 samples of tap water were taken from the user in a highly industrialized area using the Gao-Ping River as raw water.Because of the very low levels of PCDD/Fs present in water, an in situ pre-concentration system was used for sampling of source water (averagely 554 L, n = 17) and tap water (averagely 1360 L, n = 18).This system consisted of a pump which could be set in the range of 0.2-0.8L/min, a glass fiber filter holder, polyurethane foam holders and a water meter to measure the volume of water sampled.

PCDD/Fs in the Ambient Air of Drinking Water Treatment Plant LN
In 2003, a total of 4 samples of ambient air were taken from the atmosphere of DWTP-LN simultaneously with the water sampling.Each ambient air sample was collected using a PS-1 sampler (Graseby Anderson, GA, USA) according to the US EPA Compendium Method TO-9A.The sampling flow rate was 0.225 m 3 /min.Each sample was collected continuously on three consecutive days, yielding a sampling volume of 972 m 3 .The PS-1 sampler was equipped with a quartz-fiber filter for sampling particle-phase PCDD/Fs, which was followed by a glass cartridge containing PUF for sampling gas-phase PCDD/Fs.A known amount of surrogate standard was spiked to the PUF in the laboratory before the field sampling was conducted.The recoveries of PCDD/F surrogate standards ranged between 82% and 115%.They met the criteria within 70-130%.Relative standard deviation also calculated and presented in relevant tables.
Considering that PCDD/F concentrations in the ambient air generally vary with different seasons, the sampling and analyzing of these four samples were scheduled in spring, summer, fall, and winter, respectively.Furthermore, based on the variation of meteorological conditions (such as precipitation, sunshine, temperature, etc.) near the sampling sites in the past few years, representative days were thus selected for ambient air sampling.

Analysis of PCDD/Fs
PCDD/F analyses of water and ambient air samples followed the US EPA Method-1613B and EPA Compendium Method TO-9A, respectively.All chemical analyses were conducted by the Super Micro Mass Research and Technology Center in Cheng-Shiu University, certified by the Taiwan EPA for analyzing PCDD/Fs.Each collected sample was spiked with a known amount of the internal standard solution to the extraction thimble prior to PCDD/F analysis.After being extracted for 24 h, the extract was concentrated, treated with concentrated sulphuric acid, and subjected to a series of sample cleanup and fractionation procedures.The elute was concentrated to approximately 1 mL and transferred to a vial.The concentrate was further concentrated to near dryness, using a stream of nitrogen.Immediately prior to analysis, the standard solution was added to the sample to ensure recovery during the analysis process.A high-resolution gas chromatographs/high-resolution mass spectrometers (HRGC/HGMS) were used for PCDD/F analyses.The HRGC (Hewlett Packard 6970 Series, CA, USA) was equipped with splitless injection and a DB-5 fused silica capillary column (60 m length, 0.25 mm ID, and 0.25 m film thickness) (J&W Scientific, CA, USA).The oven temperature program was set as follows: begin at 150°C (held for 1 min), then increase by 30°C/min to 220°C (held for 12 min), then increase by 1.5°C/min to 240°C (held for 5 min), and finally increase by 1.5°C/min to 310°C (held for 20 min).Helium was used as the carrier gas.The HRMS (Micromass Autospec Ultima, Manchester, UK) was equipped with a positive electron impact (EI+) source.The analyzer mode of the selected ion monitoring (SIM) with a resolving power at 10,000 was used.The electron energy and source temperature were specified at 35 eV and 250°C, respectively.
For analysis of PCDD/Fs, blank tests were implemented for both filter and PUF.One set of blanks, one for filter and one for PUF, were conducted for every ten actual samples.Field and laboratory blank samples were routinely analyzed for quality assurance purposes.However, the PCDD/F levels in blank tests were normally very low and not detectable in the present study.Thus blank correction was not required.For the samples collected from the ambient air, the method detection limits (MDLs) conducted in this study were 0.8 fg/m 3 for OCDD and < 0.1 fg/m 3 for other congeners; While for the samples collected from source and tap water, MDLs were 0.2 fg/L for OCDD and < 0.01 fg/L for other congeners.

Atmospheric Deposition of PCDD/Fs on the Water Surface of Treatment Facilities
Using the measured data of PCDD/Fs in the atmosphere of DWTP-LN and theoretical model, the gas/particle partition, gas/particle deposition velocity, and dry deposition flux were built up.For wet deposition, we used the gas-and particle-phase concentrations, gas scavenging ratio, and particle scavenging ratio, then the dissolved and particle phase concentrations of rain water can be obtained.Finally by taking into account the weather information gathered from the nearest weather station, the annual dry and wet deposition can be obtained.
Atmospheric deposition of PCDD/Fs on the water surface of drinking water treatment facilities will mix with the water, and a fraction of PCDD/F mass in water will be removed by a conventional treatment process.The increased concentration of PCDD/Fs in water via deposition will be compared to the level in source water and tap water.

PCDD/Fs in Source Water
As shown in Table 1, the means of total PCDD/F concentrations in source water are 0.0049, 0.0086, and 0.0107 pg WHO-TEQ/L for Sin-Shan Reservoir, Jhuo-Shuei River, and Gao-Ping River, respectively.They are all much lower than the environmental water quality standard (1 pg WHO-TEQ/L) issued by the government of Japan, indicating that Taiwan government has paid much attention to improve environmental water quality.Fig. 1 shows the congener profiles of 17 PCDD/Fs in three different source water.The mass fractions of 17 PCDD/Fs in three source water are very similar, and the top three predominant congeners are OCDD, 1,2,3,4,6,7,8-HpCDD and OCDF.

PCDD/Fs in Tap Water
Table 2 shows the mean concentrations of PCDD/Fs in tap water.The mean concentrations of total PCDD/Fs are 0.0039, 0.0013, and 0.015 pg WHO-TEQ/L for LN, GS, and RW area, respectively.They are all lower than those in the raw water before treatment and are also lower than the environmental quality standard for dioxins in drinking water (12 pg WHO-TEQ/L) proposed by Taiwan EPA (http://tsm.epa.gov.tw/drinkwater/law/law04.htm);Furthermore, they are even lower than the current maximum allowable dioxins level of 1 pg WHO-TEQ/L in Japan (Environmental Health Department, 2002).These results reveal that the purification units in the treatment plant could effectively removed PCDD/Fs.Fig. 2 shows the congener profiles of 17 PCDD/Fs in tap water.Comparing with the source water, the most predominant congener is still OCDD (Fig. 1 and 2), but the patterns in tap water (Fig. 2) are somewhat different from those in the source water (Fig. 1).
As mentioned in the METHOD part, the water source of LN area is Jhuo-Shuei River, and those of GS and RW areas are Gao-Ping River.Thus, the PCDD/F levels of tap water from LN, GS and RW area and those of the source water (Sin-Shan reservoir, Jhuo-Shuei River and Gao-Ping River) are not comparable.

PCDD/Fs in the Ambient Air of Drinking Water Treatment Plant LN
Table 3 shows the concentrations of PCDD/Fs in the ambient air of DWTP-LN.The concentrations of total PCDD/Fs ranged from 0.024 to 0.493 pg I-TEQ/Nm 3 with an average of 0.277 pg I-TEQ/Nm 3 , which is approximately three times (0.277/0.088) higher than that in other area (0.088 pg I-TEQ/Nm 3 ) in Taiwan (Wang et al., 2005).Fig. 3 shows the congener profiles of 17 PCDD/Fs for the air of DWTP-LN.The predominant congeners are OCDD, followed by 1,2,3,4,6,7,1,2,3,4,6,7,and OCDF.The congener profile of ambient air was different from stack flue gas from a municipal solid waste incinerator or from unburned joss paper, such as OCDD, HpCDF, HpCDD, HxCDF and HxCDD (Hu et al., 2009;Wang et al., 2009).

Atmospheric Deposition of PCDD/Fs on the Water Surface of Treatment Facilities Particle/gas partitions
An equation that has been used with success to describe particle-gas partitioning is as follow: Where K p (m 3 / g) is a temperature-dependent partitioning constant, TSP ( g/m 3 ) is the concentration of total suspended particulate material, F (pg/m 3 ) is the concentration of the  (Yamasaki et al., 1982;Pankow, 1991;Pankow and Bidleman, 1992;Pankow, 1994).Plotting log K p against the logarithm of the subcooled liquid vapor pressure, Where m r is the slope and b r is the y-intercept of the trendline (Lohmann and Jones, 1998).Eitzer and Hites (1988) have correlated P L o of PCDD/Fs with gas chromatographic retention indexes (GC-RI) on a non-polar (DB-5) GC-column using p,p'-DDT as a reference standard.The correlation has been redeveloped by Hung et al. (2002) log P L o = -1.34(RI)/T + 1.67 × 10 -3 (RI) -1320/T + 8.087 (3) Where P L o is subcooled liquid vapor pressure, RI is gas chromatographic retention indexes derived by Donnelly et al. (1987) and Hale et al. (1985), and T is ambient temperature (K).In this study, the RIs derived by Donnelly et al. (1987) and Hale et al. (1985) and Eq.(3) redeveloped by Hung et al. (2002) were taken to generate the P L o values.A complete dataset on the gas-particle partitioning of PCDD/Fs in Taiwan has been reported (Chao et al., 2004).The data gave values for m r = -1.29 and b r = -7.2 with R 2 = 0.94.In this study, the trendline proposed by Chao et al. (2004) was taken to estimate the partitioning constant, Kp.
Table 4 lists the environmental conditions during the sampling campaign.The temperature ranged from 18.1 to 29.5°C with an average of 23.5°C, and the concentration of total suspended particulate material (TSP) ranged from 86 to 327 g/m 3 , with an average of 212 g/m 3 .In years, Asian dust storm (ADS) that originated in the deserts of Mongolia and Mainland China eventually reached populated areas of East Asia, including Taiwan.Prior to the ADS episode, the atmospheric PCDD/F concentrations were considerably lower than those measured in other Asian countries.Nevertheless, they increased 1.9-3.2times during the ADS episode.Notably, the amount of PCDD/Fs bound to suspended particles increased from 257-259 to 339-512 pg I-TEQ/g TSP (Chi et al., 2008).As shown in Table 4 of the present study, the significantly higher TSP The PCDD/Fs bound to suspended particles would thus increase and then led to the relatively higher PCDD/F concentrations in the ambient air, particularly during winter (0.493 pg I-TEQ/Nm 3 in Table 3).Additionally, the lower temperatures during fall and winter seasons (Table 4) tend to lower the altitude of boundary layer.Within this layer, temperature increases with altitude and result in poor diffusion of air pollutants.As a result, the higher PCDD/F concentrations were measured in the ambient air.
Based on the environmental conditions, the calculated sub-cooled liquid vapor pressure (P L o ) and gas-particle partitioning constant (K p ) of PCDD/Fs in the ambient air of DWTP-LN during the four sampling periods are shown in Table 5.This result shows that a compound with lower chlorine numbers has a higher sub-cooled liquid vapor pressure.It also illustrates that a compound with higher chlorine number has a higher particle-gas partitioning constant.So, at constant temperature and TSP concentrations, the congeners of PCDD/Fs with a higher chlorine number will have a higher fraction of particle phase.
The percentage of particle-gas partitioning in the ambient air of DWTP-LN is shown in Table 6.The fraction of the gas phase for congener 2,3,7,8-TeCDD ranged from 0.72 to 0.97, and 0.79 to 0.98 for 2,3,7,8-TeCDF.For the congeners with the highest chlorine number, the fraction of the particle phase for congener OCDD ranged from 0.95 to 1.00 and 0.93 to 1.00 for OCDF.

Atmospheric dry deposition of PCDD/Fs
The atmospheric dry deposition flux of PCDD/Fs is a combination of both gas-and particle-phase flux, which is given by where F T is the total PCDD/F deposition flux contributed by the summation of both gas-and particle-phase flux, F g is the PCDD/F deposition flux contributed by the gas phase, F p is the PCDD/F deposition flux contributed by the particle phase, C T is the measured concentration of total PCDD/Fs in the ambient air, V d,T is the dry deposition

Sin-Shan Reservoir n=1
Fraction (%) velocity of total PCDD/Fs, C g is the calculated concentration of PCDD/Fs in the gas phase, V d,g is the dry deposition velocity of the gas-phase PCDD/Fs, C p is the calculated concentration of PCDD/Fs in the particle phase, and V d,p is the dry deposition velocity of the particle-phase PCDD/Fs.The mean dry deposition velocity of total PCDD/Fs (0.42 cm/s) was proposed by Shih et al. (2006).This value (V d,T = 0.42 cm/s) is also used for the approximate calculation of total PCDD/F dry deposition flux.
Dry deposition of gas-phase PCDD/Fs is mainly by diffusion, because of a lack of measured data for PCDD/Fs, a selected value (0.010 cm/s) of gas-phase PAH dry deposition velocity, V d,g , proposed by Sheu et al. (1996) and used by Lee et al. (1996) is also used here to calculate the  1 , 2 , 3 ,4 ,7 ,8 -H x C D F 1 , 2 , 3 ,6 ,7 ,8 -H x C D F 1 , 2 , 3 ,7 ,8 ,9 -H x C D F 2 , 3 , 4 ,6 ,7 ,8 -H x C D F 1 , 2 , 3 ,4 ,6 , 7 , 8 -H p C D F 1 , 2 , 3 ,4 ,7 , 8 , 9  PCDD/F dry deposition flux contributed by its gas phase.Dry deposition of particle-phase PCDD/Fs is mainly achieved by gravitational settling, and the dry deposition velocity of particle-phase PCDD/Fs, V d,p , can be calculated by Eq. ( 4).The calculated dry deposition velocity of particle-phase PCDD/Fs, and the data bank required for this calculation, are listed in Table 7.The data shows that the mean dry deposition velocity of particle-phase PCDD/Fs is 0.44 cm/s for the ambient air of DWTP-LN.

LN area n=4
Table 8 lists the deposition fluxes of PCDD/Fs in the ambient air of DWTP-LN.The mean dry deposition fluxes of PCDD/Fs contributed by the gas phase and particle phase are 0.769 and 71.5 pg/m 2 -day, respectively, and 98.9% was contributed by particle-phase deposition.These results demonstrate that the dry deposition of PCDD/Fs was  primarily contributed by the particle phase.This is due to that the dry deposition velocity of particle-phase PCDD/Fs (0.44 cm/s) was much higher than that of gas-phase velocity (0.010 cm/s).

Atmospheric wet deposition
The wet deposition flux of PCDD/Fs was a combination of both vapor dissolution into rain and the removal of suspended particulate by precipitation.For a slightly soluble trace organic compound as PCDD/Fs, it is commonly thought that equilibrium partitioning occurs between the compound in the gas phase and a falling rain drop (Ligocki et al., 1985a, b).The gas scavenging ratio, S g , can be estimated by Where S g is the gas scavenging ratio of PCDD/Fs (dimensionless), R is the universal gas constant (82.06 × 10 -6 m 3 atm/mol-K), T is ambient temperature (K), and H is the Henry constant (m 3 atm/mol).On the other hand, particle scavenging is not based on equilibrium considerations and depends largely on meteorological factors and particle characteristics.Scavenging ratio is defined as the concentration of the pollutant in the raindrop divided by the concentration in the surrounding air during the precipitation event.The gas scavenging ratio, S g , can be calculated by where S g is the gas scavenging ratio of PCDD/Fs (dimensionless), it is a ratio of the concentration of the dissolved phase in the raindrop divided by the concentrations of the gas phase in the air, C rain,dis is the   (Donnelly et al., 1987;Hale et al., 1985) b log P L o = -1.34(RI)/T + 1.67 × 10 -3 (RI) -1320/T + 8.087 (Hung et al., 2002)  c Using the equation of : log K p = -1.29 log P L o -7.2 (Chao et al., 2004) Table 6.Particle-gas partitioning in the ambient air of drinking water treatment plant LN.
Sampling periods 3/ 25-3/28, 2003 6/30-7/3, 2003 11/18-4/21, 2003  Where S p is the particle scavenging ratio of PCDD/Fs (dimensionless), it is a ratio of the concentration of the particle phase in the raindrop divided by the concentrations of the particle phase in the air, C rain,particle , is the particle-phase concentration of PCDD/Fs in the raindrop, and C p is the concentration of PCDD/Fs in the particle phase.
Total scavenging of precipitation (S tot ) is the sum of gas and particle scavenging, it can be calculated by Where S tot is the total scavenging ratio of PCDD/Fs (dimensionless), and is the fraction of the total air concentration bound to particles.
The particle scavenging ratio of PCDD/Fs is based on the average Bloomington air and rain concentrations that were measured by Eitzer and Hites (1989).Suppose the seventeen 2,3,7,8-substitued PCDD/Fs were evenly distributed among different particle sizes and the particles in ambient air were wash out uniformly, a common value should be acquired for all congeners.But the data measured by Eitzer and Hites (1989) exhibit a wide diversity of S p values.There is a possibility that a congener dissolved in a raindrop is re-adsorbed by the particle scavenged.Accordingly, the S p values of congeners which have lowest solubility for PCDD and PCDF, i.e., OCDD and OCDF, should be more accurate.Because of a lack of measured data, the S p values of OCDD and OCDF measured by Eitzer and Hites (1989) were averaged and used here.
The scavenging ratios of PCDD/Fs in the ambient air of DWTP-LN are shown in Table 9.Among the 7 congeners of PCDDs, S g values ranged from 5.46 × 10 2 (HxCDD) to 9.35 × 10 3 (PeCDD).For the 10 congeners of PCDFs, S g values ranged from 1.65 × 10 3 (TeCDF) to 2.48 × 10 4 (OCDF).There is no consistent trend between S g values and degree of chlorination.The average particle scavenging ratio of OCDD and OCDF measured by Eitzer and Hites (1989), i.e., 42,000, is used in this study.
Based on gas and particle scavenging ratios, the dissolved and particle phase concentrations of raindrops, C rain,dis and C rain,particle , can be calculated by Eqs. ( 6) and ( 7), respectively.Table 10 lists the calculated raindrop concentrations of PCDD/Fs at DWTP-LN.At WTP-LN, the mean dissolved and particle phase raindrop concentrations of PCDD/Fs are 1.65 and 438 pg/L, respectively, and the particle phase contributed 99.6% of those.The dissolved and particle phase raindrop TEQ concentrations of PCDD/Fs are 0.367 and 7.90 pg I-TEQ/L, respectively, and the particle phase contributed 95.6% of those.

Annual Dry and Wet Deposition Flux of PCDD/Fs
In order to estimate the annual deposition flux of PCDD/Fs, the information of precipitation and days of precipitation in 2003 were gathered from Jia-Yi weather station, and are listed in Table 11.The annual precipitation is 941 mm, and the annual days of precipitation are 60 days.The precipitation was not even through the year, it was concentrated in August to September.The annual deposition flux of PCDD/Fs entering the DWTP-LN is listed in Table 12.The total deposition flux of PCDD/Fs is 1439 ng/m 2 -year, and wet deposition contributed 10.0% of those.With regard to TEQ, the total deposition flux of PCDD/Fs is 27.0 ng I-TEQ/m 2 -year, and wet deposition contributed 12.0% of those.The above results reveal that dry deposition is more important than wet deposition for total deposition flux of PCDD/Fs in the ambient air of DWTP-LN.Comparison with similar semi volatile organic compounds, polycyclic aromatic hydrocarbons (PAHs), the dry deposition is the dominant disappearance mechanism during sampling period and also the high dry deposition velocities are threatening to human health (Chang et al., 2003;Sahu et al., 2008).

Atmospheric Deposition of PCDD/Fs on Water Surface of Treatment Facilities
The treatment capacity of DWTP-LN is 198,000 m 3 /day, with a water surface area of 23,840 m 2 to accept deposition from the atmosphere.According to the result described previously, there was a deposition flux of 27.0 ng I-TEQ/m 2 -year, and it will mix with the water treated.If the deposition was not removed by the treatment units, it will increase the PCDD/F concentration in water up to 8.91 × 10 -3 pg I-TEQ/L, which is about 92% of PCDD/Fs in raw water.Kim et al. (2002) reported that the majority of PCDD/F mass are well removed (removal efficiency 87%) by conventional rapid sand filtration, thus the treated water may increase 1.16 × 10 -3 pg I-TEQ/L, which is about 43% of the level in tap water.The above results indicate that the influence of atmospheric deposition of PCDD/Fs on drinking water is of great significance and the water treatment facility with a cover is suggested.

CONCLUSIONS
The results of this study come to the following conclusions: (1) The concentrations of PCDD/Fs in source water ranged from 0.0046 to 0.0110 pg I-TEQ/L with an average of 0.0084 pg I-TEQ/L.(2) The concentrations of PCDD/Fs in tap water ranged from 0.0012 to 0.0037 pg I-TEQ/L with an average of 0.0021 pg I-TEQ/L, which is much smaller than that in source water.(3) The total atmospheric deposition flux of PCDD/Fs entering the DWTP-LN was 27.0 ng I-TEQ/m 2 -year, and wet and dry deposition contributed approximately 12.0% and 88%, respectively.It reveals that dry deposition is more important than wet deposition.(4) Atmospheric deposition of PCDD/Fs will increase the level in tap water up to 8.91 × 10 -3 pg I-TEQ/L, which is approximately 92% of that in source water.If a removal efficiency of 87% is achieved by conventional rapid sand filtration system, the tap water may increase 1.16 × 10 -3 pg I-TEQ/L, which is about 43% of the level in tap water.These results indicate that in the DWTP-LN, the influence of atmospheric deposition of PCDD/Fs on the drinking water is of great significance, and supplying water treatment facilities with a cover is suggested.

Fig. 3 .
Fig. 3. Congener profiles of 17 PCDD/Fs in the ambient air of drinking water treatment plant LN.

Table 1 .
PCDD/Fs in source water

Table 2 .
PCDD/F levels in tap water

Table 3 .
PCDD/F concentrations in the ambient air of DWTP-LN.

Table 4 .
Environmental conditions during the sampling campaign.

Table 5 .
Sub-cooled vapor pressure and particle-gas partitioning constant of 17 PCDD/Fs in the ambient air of WTM-LN.

Table 11 .
Precipitation and days of precipitation in 2003 in LN area.