210Po Released from the Iron- and Steel-Making Plant ILVA in Taranto (Italy) on the Environment and the Public

Lead-210 and Po are naturally occurring radionuclides. Due to volatile characteristic of lead and polonium, environmental pollution of Pb and Po released from the coal power plant, steel-making industry and refractory material industry has been an exposure problem for the members of public. In this paper studies on the activity concentrations of Po and Pb in the raw materials, dust particles, surficial soils and atmospheric particulate samples collected in the area of the Ironand Steel-Making Plant ILVA Taranto (Italy) were made. These data have been used to evaluate the source-term, distributions, inventories, mass balance, biological availability, ecological migration processes and public exposure risk of Pb and Po in the concerned environment.


Introduction
In general, the main source of 210 Pb and 210 Po in the environment is the exhalation of 222 Rn gas from the ground into the atmosphere. However, as a result of (i) the volatile characteristic of lead and polonium elements and (ii) the development of industries in recent decades, that is, mining, processing, and smelting of uranium, phosphate, lead and iron ore, burning of fossil fuels (coal), and burning leaded gasoline used for car engines during transportation, elevated activities of 210 Pb and 210 Po have been found in the atmosphere, hydrosphere, biosphere and surficial soil. A numerous of blood/food/dust/soil lead contamination cases have been reported [1][2][3][4]. Therefore, the artificial contamination of 210 Po, 210 Pb, and stable lead is more and more seriously affecting the public health via air inhalation and food ingestion.
Some years ago, Italian researchers started the contamination source-term survey on the natural occurring radionuclides, including uranium isotopes, thorium isotopes, radium isotopes, 210 Pb, and 210 Po [5]. The purpose of the project was mainly focused on studies of the contamination process, the exposure risk evaluation to the public, and the remedial measures for radiation protection from the radionuclides released from the coal power plant, steel-making industry, and refractory material industry. The obtained results (Table 1) [5] showed that to lower the exposure risk to the public great attention should be given not only to the process of raw material supply and final product utility, but also to the process of by-product redistribution, high enriched waste management, and disposal of the natural occurring radionuclides in the studied industries, especially of the uranium isotopes, thorium isotopes, radium isotopes, 210 Pb and 210 Po, and so forth.
Following the warning by a press about the possible contamination of 210 Pb and 210 Po emitted from the chimney of the Iron-and Steel-Making Plant ILVA in Taranto (Italy) into the atmosphere, in October 2008, the Italian Ministry of Environment, Territory and Sea Protection asked the National Institute of Environmental Protection and Research (ISPRA) [6,7] to prepare a detailed report to evaluate the radiological risk of the radioactive pollution arising from the plant itself. For this purpose, the ISPRA developed a program for the information acquisition, the contaminants and exposure pathway analyses, the realization of the contaminant monitoring, and the radiological impact evaluation. In the process of implementation of the program, the raw materials used in the plant, soil samples surrounding the plant, dust/ash

The Study Area and
Sampling. Taranto city is located at a latitude (N) of 40 ∘ 28 00 and a longitude (E) of 17 ∘ 14 00 in Puglia region, the far south of Italy, with an area of 210 km 2 , 15 m above sea level, and population of 192 thousands. Meteorological data of the region collected in 30 years show that the mean annual precipitation, and minimum and maximum temperatures were 416.5 mm, 12.7 (6.0-20.9) ∘ C, and 20.5 (12.2-29.9) ∘ C, respectively.
The Iron-and Steel-making Plant ILVA in Taranto, just situated in the north of Taranto city and owned by the Riva Group, is the ILVA's core integrated iron-and steel-making plant and the largest plant in Europe. Production capacity is around 12 Mt of steel per year. The plant comprises five blast furnaces, AFO 1, 2, 3, 4, and 5. The iron mineral (85%), limestone (3-16%), coal (or coke: 3.5%), and lime (0.4-0.8%) are used as the principal raw materials in the plant. These materials are charged in batches into the blast furnaces at temperature of over 1000 ∘ C where the iron compounds in the ore give up excess oxygen and become liquid iron. At intervals of a few hours, the accumulated liquid iron is tapped from the blast furnace and either cast into pig iron or directed to other vessels for further steel-making operations.
Due to the fact that the raw materials always contain volatile materials, this process of iron-or steel-making has caused the environmental pollution of nitrogen oxides, sulphides, arsenide, carbon dioxide, organic compounds, heavy metals, naturally occurring radionuclides, and so forth in air, water, and soil or sediment. It was reported that in 2002 the ILVA in Taranto emitted 30.6% of total dioxin emission in Italy [8], and based on the data from National Emission Inventories and Sources (INES) the percentage of the dioxin emission from the ILVA in Taranto raised to 92% of the total dioxin emission in Italy in 2006. Mercury contamination from the ILVA in Taranto is another important environmental problem, and it was estimated that in 2005 the annual total mercury emission was over 2 t.
As far as naturally occurring radionuclides are concerned, attention should be given to uranium isotopes, thorium isotopes, radium isotopes and progenies (e.g., 210 Pb and 210 Po), and so forth. Especially, the relatively high contaminations of 210 Pb and 210 Po in air, water, and soil are often predicted to happen, due to the volatile behaviour of lead and polonium at the temperature >1000 ∘ C. However, there is little information about the contamination level and exposure risk evaluation of 210 Pb and 210 Po released from the ILVA in Taranto.
For better understanding the source-term and inventory of 210 Pb and 210 Po in production process of the ILVA in Taranto, raw materials used in the plant, including lime, limestone, coal (coke), and iron minerals, were collected on May 5, 2009. For information about the redistribution and contamination level of 210 Pb and 210 Po in the environment, fly ash (dust particle) deposited in the filtration systems and dust particles emitted from the E312 chimney (latitude: 40 ∘ 28 00 N, longitude: 17 ∘ 14 00 E) were collected on May 5-7 2009. For the purpose of the exposure impact evaluation for the members of public, atmospheric particulate samples were collected by ISPRA in the sampling campaign of November 2008 and by ARPA Puglia in the sampling campaign May 2009. In the early days of November 2008 seven surficial soil samples were collected, in which five samples were taken in the areas around the ILVA in Taranto, one in Gioia del Colle about thirty miles northwest of Taranto and one in Castel Romano in Roma (CSM Roma). The sampling depth (0-5 cm) of all seven soil samples was the same. The sampling sites for air and soil samples were selected preferring areas of probable increased exposure on the basis of wind prevalence. Castel Romano in Roma being the Laboratory of the ISPRA was selected as a control site for both air and soil samples for comparison. The detailed information about the sampling site, strategy, date, and parameters for atmospheric particulate and soil was given in Tables 2, 3, and 4, respectively.

Apparatus and Reagents.
Bismuth-210 for 210 Pb determination was measured by a 10-channel low-level -counter (Berthold LB770, Germany). The instrument and reagent background of the counter for 210 Pb measurement is of ≤0.0053 cps, and the counting efficiency was 48.2% that was calibrated with a PbSO 4 precipitate source obtained from a standard 210 Pb solution. Po-210 was determined by alpha spectrometry (Canberra, USA) with a counting efficiency of 31.2% and a background of ≤ 6 × 10 −6 cps in the interested energy region.
A Perspex disk holder for polonium deposition was specially designed to fit 100-250 mL beakers [9]. Silver foil with a thickness of 0.15 mm was used for 210 Po spontaneous deposition, and it was cut into disks of 23 mm in diameter. Large volume (67 m 3 h −1 ) air sampler was the Model Thermo G10557 equipped with analyzer of PM 2.5 or PM 10 and with a glass-microfiber filter of dimensions of 20 cm × 25 cm (Whatman GF/A cat. N. 1820-866).
Polonium-209 solution standard as a tracer for 210 Po determination by -spectrometry and 210 Pb solution standard for -instrument calibration, the reference material (IAEA-315) for quality control, and the BIO-RAD-AG 1-X4 resin (100-200 mesh) for lead separation were supplied by Amersham (UK), the IAEA, and the Bio-Rad Laboratories (Canada), respectively. TOPO (tri-octyl-phosphine oxide, 99%) used to isolate 210 Po, Pb(NO 3 ) 2 to prepare the carrier solution for lead separation, pure iron wire to prepare standard iron solution for chemical yield calculation and all other reagents were analytical grade.  Pb in the normal soil samples were analyzed following the procedures reported in a literature [10]. The filtered air samples was composed of glass-fiber filter paper and dust/ash or the suspended particle from air. The analytical procedure for the air filter samples were as the same as that for normal soil samples, except for avoiding addition of 40% HF during leaching. Some samples, such as the raw materials, dust particles collected in the filtration systems, fly-ash, and so forth, showed a very high iron contents, and serious interferences for both 210 Pb and 210 Po determinations had Two mL of 0.5 M Na 2 S was added, and in this case PbS was precipitated while the remaining Ca 2+ and Mg 2+ will remain in the solution. After centrifugation, the supernatant was discarded and the black precipitate was dissolved with 3 mL of concentrated HCl and 21 mL of distilled water by heating. The obtained solution was passed through a preconditioned anion-exchange resin column at room temperature and at a free flow rate. After washing with 40 mL of 1.5 M HCl, lead was eluted with 60 mL of distilled water at free flow rate, and the separation time of the pair 210 Pb/ 210 Bi was recorded. Two mL of conc. H 2 SO 4 was added to the collected eluant, which was then evaporated until fuming to destroy the organic matters by oxidation with 1 mL of 30% H 2 O 2 . Both the precipitate and the solution were centrifuged. The supernatant was discarded, and the precipitate was filtered on a weighed filter paper with a diameter of 24 mm (Whatman 42). The filter was dried at 110 ∘ C until constant weight (about 1 h) and weighed again to calculate the lead chemical yield. Lead-210 was determined by measuring the ingrowth activity of its progeny 210 Bi (T 1/2 : 120 h) by a low background -counter some time after the separation (about one month of being suitable). The 210 Pb activity concentration ( Pb-210 ) in soil sample (Bq kg −1 ) or in air sample (Bq m −3 ) was calculated according to the following equation: where Bi−210 is the net count rate of 210 Bi (cps); Bi , the 210 Bi decay constant (min −1 ); , is the 210 Bi ingrowth time after 210 Pb separation (min); , is the detection efficiency for 210 Bi; 6 Journal of Chemistry , is the chemical yield; , is the sample weight (kg) for soil or the volume (m 3 ) for air.
Quality Control. Following approaches can be used to review the quality of a radioanalytical method: (1) to analyze the certified reference materials or similar matrices and to compare the obtained results with the recommended values, (2) to participate in the intercomparison activities between different international laboratories, and (3) to analyze the spiked samples.
For the purpose of quality control, the reference material IAEA-315 Marine Sediment supplied by the IAEA was used, in which the recommended value of 210 Pb was given. About 2 g of the reference material was analyzed following the recommended procedure of this paper. The precision was evaluated by the relative standard deviation obtained from a set of six analyses. The accuracy was assessed by the term of relative bias, which reflects the difference between the experimental mean and recommended value of 210 Pb activity concentration. Due to the presence of unsupported 210 Pb in the IAEA-315, the fraction of unsupported 210 Pb had to be corrected to the base date.
The obtained 210 Pb activity concentrations in the IAEA-315 were shown in Table 5. The mean 210 Pb concentration in the IAEA-315 was found to be 30.7 ± 1.7 Bq kg −1 (decay correction to the date of 1st January 1993). It was observed that the relative standard deviation is ±5.5% for 210 Pb. Since all being less than ±10% the precision for the analyses is well accepted as far as such a low activity is concerned. The relative bias obtained from the analyses was +2.0% for 210 Pb, showing that the mean activity concentrations of 210 Pb are in good agreement with the recommended value of 30.1 Bq kg −1 (the 95% confidence interval: 26.0-33.7 Bq kg −1 ).
Due to its short half life, the reference materials for 210 Po are not available. The quality control for 210 Po analyses in this laboratory was carried out through participating in the intercomparison activities organized by the IAEA in March 29, 2007. The samples for intercomparison were a set of five water samples. The obtained activity concentrations of 210 Po were all in good agreement with the values given by the IAEA.
Detection Limits. Taking into account the 3 of the blank count rates, the counting efficiencies of the instrument, the radiochemical yields, the ingrowth or decay factor ( 210 Pb: 100%) and the sample weight or volume, and the detection limit, or more precisely the minimum detectable activity (MDA) of the method for soil and air samples, are 0.25 Bq kg −1 and 1.7 Bq m −3 for 210 Po and 0.73 Bq kg −1 and 1.7 Bq m −3 for 210 Pb, respectively.

Determination of Lead in Dust Particle
Samples Taken from Chimney. The procedure for stable lead separation and determination in the dust particle samples taken directly from chimney was the same as that for 210 Pb in soil samples, except for not adding lead carrier. The chemical yield for such dust particle sample determination was obtained through additional analyses of 16 soil samples without artificial lead contamination by addition of lead carrier. The lead concentration was obtained from the PbSO 4 weight of the sample after weighing and chemical yield corrections. Taking into account the weighing deviation, chemical yield, and sample quantity, the estimated minimum detectable quantity for lead was 0.0032%.

Determination of Iron Concentration in Raw Materials
Determination of the Iron Chemical Yield. The standard iron solution (40.0 mg Fe mL −1 ) was prepared by dissolution of pure iron wire with 6 M HCl and some 30% H 2 O 2. One mL of the standard iron solution was put in a beaker, and the further treatment was done following the procedure title (Iron Separation). The iron chemical yield was calculated as the ratio of Fe weight in the residue over that in the standard solution taken. Four-time repeated analysis showed that the iron chemical yield of the procedure was 99.4 ± 2.2%.
Leaching. Three g of raw material, 20 mL of conc. HCl, and 10 mL of 40% HF were added to an 100 mL Teflon beaker. The beaker was heated at 250 ∘ C. Before drying, 20 mL of conc. HCl and 10 mL of 40% HF each were added, and the step was repeated. The solution was evaporated to incipient dryness and a 10 mL portion of 72% HClO 4 was added. The solution was evaporated to fuming to destroy the organic matter and to remove HF. The HClO 4 treatment was repeated until nearly the entire solid sample was decomposed. The residue was finally dissolved with 15 mL of conc. HCl and 40 mL of water. The obtained solution was filtered through a 0.1 m Millipore filter paper.
Iron Separation. A portion of the leaching solution was put in a beaker, and some conc. ammonia solution was added to precipitate iron as Fe(OH) 3 at pH 9-10. After heating to flocculate the precipitate well, the sample was transferred to a plastic centrifugation tube and centrifuged at 4000 rpm for 5 min. The supernatant was discarded, and the iron precipitate was dissolved with 20-30 mL of conc. HCl. The obtained solution was passed through a preconditioned anion-exchange resin column at room temperature and at a free flow rate [11]. After washing with 5 mL of 9 M HCl, 30 mL of 6 M HCl + 1% H 2 O 2 , and 30 mL of 4 M HCl, iron Calculation of the Iron Concentration. The Fe 2 O 3 weight was corrected by the iron composition, the sample weight and chemical yield, and the iron concentration (%) in the sample was obtained. Taking into account the weighing deviation, chemical yield, and sample quantity, the estimated minimum detectable quantity for iron was 0.0030%.

Results and Discussion
Tables 5-11 show the activity concentrations of 210 Po and 210 Pb in the analyzed samples. The given uncertainties (1SD) were estimated taking into account the errors associated with the weighing samples, instrument calibration, yield calculation and the counting statistics of the sample and the blank sources, and so forth.

The 210 Po and 210 Pb Concentrations in Raw Material
Samples. The raw material is the unique source of 210 Po and 210 Pb in the plant ILVA in Taranto. The contamination of 210 Po and 210 Pb in air is proportional to their volatile characteristics, to the activity concentrations of the 210 Po and 210 Pb in raw materials, and to the productivity of iron or steel and is inversely proportional to their adsorption characteristics on particle before filtration and to the filtration efficiency of the filtration system. The main raw materials used in the iron-/steel-making process are iron ore, and coke as a reductant and limestone and lime as flux to make a blast furnace more efficient. As shown in Table 6, the activity concentrations of 210 Po and 210 Pb in coke were low if compared with their concentrations in soil for Europe (a few Bq kg −1 up to hundreds of Bq kg −1 ) [12]. It is not a surprise as the preparation process of coke could let the major part of 210 Po and some parts of 210 Pb in coal volatilize. The data in Table 6 showed that some 210 Pb still remain in coke, and the secular equilibrium between 210 Pb and 210 Po has not been reached with a 210 Po/ 210 Pb ratio of only 0.819. The activity concentrations of 210 Po and 210 Pb in limestone, lime and iron minerals were also low, especially in MIN 3, but 210 Po and 210 Pb seem to be in secular equilibrium with 210 Po/ 210 Pb ratios of ∼1 in these samples. However, The activity concentrations of 210 Po and 210 Pb in homogenized and mixture of agglomeration samples were similar and high if compared with those in the raw materials, and it was intimated that except for coke, limestone, lime, and iron minerals as raw materials there could be some other ingredients for the purpose of improving the characteristics of the iron/steel products, such as dolomite, quarts, phosphates, and so forth, which could contain much higher activities of 210 Po and 210 Pb.

The 210
Po and 210 Pb Inventory Estimation. The plant ILVA in Taranto comprises five blast furnaces. The iron minerals, limestone, lime, and coal (or coke) were used as the principal raw materials in the plant with a production capacity of around 12 Mt of steel per year. The obtained mean iron concentration in the mixture of agglomeration and homogenized samples was 55.4 ± 3.8%. From Table 6, it was seen that the iron concentrations in coke, limestone, and lime were very low and their Fe contribution to the steel-making was negligible. Therefore, from the mean iron concentrations in the mixed raw materials, it was estimated that the total quantity of the raw materials consumed in the plant was 22 Mt y −1 . Based on the mean concentrations of 210 Po (12.0 ± 4.7 Bq kg −1 ) and 210 Pb (12.3 ± 4.8 Bq kg −1 ) in the raw materials and the total quantity of the raw materials consumed per year, the annual inventories of 210 Po and 210 Pb in the area of the ILVA in Taranto could be 2.64 × 10 11 Bq and 2.71 × 10 11 Bq, respectively, if all of them were volatized from the raw materials completely, of which some deposited in the first filtration system, some in the second filtration system, and some released from the E312 chimney and deposited in the nearby environment of the ILVA in Taranto.   The volatized polonium and lead together with the flow air and the fly ash (dust particle) start to undergo a process of flowing-through/adsorption/filtration/deposition in the first (130-150 ∘ C) and second (120-140 ∘ C) filtration systems and releasing from the E312 chimney (100-130 ∘ C) and depositing in the environment of the ILVA in Taranto. At the beginning, the volatized polonium and lead could be a state of smoke, which was then step-by-step attached on dust particles or atmospheric aerosols in the submicron size range at a suitable condition. It was reported that the average attachment times were from 40 s to 3 min at environmental temperature [13]. In Pb, and stable lead in the dust particles collected in the second filtration system were higher than those in the first one. This could be explained by the fact that McNeary and Baskaran [14] proposed the hypothesis that only a small portion of the aerosols scavenges effectively 210 Pb from the atmosphere and a major portion of the aerosols do not actively participate in the removal of these nuclides from the air mass. The small portion of the aerosols should mean the portion of small particle size (<1 m). Due to the fact that the particle sizes in the first filtration system are bigger than those in the second one, the adsorption rate is lowered. DP5-DP7 in Table 8 (Table 7). From the data in Tables 7 and 8, it can be seen that the activity concentrations of 210 Po in all the DP samples except for DP3 and DP4 were much higher than those of 210 Pb. This tendency could also be explained by the melting point difference between polonium and lead elements (mp: 254 ∘ C for polonium and 327 ∘ C for lead); therefore, 210 Po is more volatile than 210 Pb. The tendency in DP3 and DP4 should also be the same; unfortunately the two samples were analyzed with more than 2 years of delay after the sampling, so the activities of 210 Pb and 210 Po nearly reached the radioactive equilibrium due to the 210 Po decay.
The activity concentrations of 210 Po or 210 Pb as a function of the stable lead contents in all dust particle samples collected in the filtration systems and at the exit of the E312 chimney of the ILVA in Taranto were shown in Figure 1. The positive correlation illustrated that the 210 Po and 210 Pb activities in the raw materials used in the plant were proportional to their lead contents.
During the investigation, the plant was operated with only 30% capacity. Based on the working condition during 9-15 April 2008, the mean daily dust particle production was 5.00 t in the first filtration system and 1.71 t in the second one. Therefore, it was estimated that the mean annual deposited quantities of the dust particles were 1826 ± 34 t y −1 in the first filtration system and 626 ± 7 t y −1 in the second one. The diameter of the E312 chimney is 10.0 m, the air flow rate was 20.0 m s −1 and the mean content of the dust particle during sampling was 5.624 mg m −3 , so the estimated dust particle quantity released from the exit of the chimney was 279 t y −1 .
According to the production quantities of the dust particles and their mean activity concentrations in the particles shown in Tables 7 and 8, the annual subinventories of 210 Po, 210 Pb, and stable lead could be 1.99 × 10 10 Bq, 1.08 × 10 10 Bq, and 7970 kg in the first filtration system, 1.41 × 10 10 Bq, 1.01 × 10 10 Bq, and 8360 kg in the second filtration system, and 2.18 × 10 10 Bq, 9.03 × 10 9 Bq, and 17758 kg released from the exit of the chimney, respectively. The total inventories estimated in this way were 5.58 × 10 10 Bq y −1 of 210 Po, 3.00 × 10 10 Bq y −1 of 210 Pb, and 34089 kg of stable lead. From this data it was seen that 23.4-36.1% of 210 Po, 210 Pb and lead were deposited in the first filtration system, 24.5-33.8% in the second filtration system, and 30.1-52.9% emitted from the E312 chimney and deposited in the environment of the ILVA in Taranto. If the plant was in operation with full capacity, the total inventory could be 1.86 × 10 11 Bq y −1 of 210 Po, 9.99 × 10 10 Bq y −1 of 210 Pb, and 1.14 × 10 5 kg of Pb. These estimated inventories of 210 Po and 210 Pb were nearly in the same order of magnitude with that (2.64 × 10 11 Bq of 210 Po and 2.71 × 10 11 Bq of 210 Pb) estimated from the raw materials in Section 3.2.  .49 times as high as those at the control site, respectively; (iv) it seems that the activity concentrations of 210 Po were a little bit less than those of 210 Pb, but they were nearly in equilibrium with a mean 210 Po/ 210 Pb ratio of 0.947 ± 0.067. The accumulation and mobility of 210 Po and 210 Pb in the terrestrial environment due to atmospheric fallout vary with the geographic locations and depend not only on physicochemical properties of the radionuclides themselves but also on (i) climatic conditions, such as rainfall, temperature, humidity, wind direction and speed, and biological activity of microorganisms in soil and (ii) human activities, such as cultivation, irrigation, and fertilization [15,16]. The deposition of 210 Po and 210 Pb from the smoke plume released from a chimney to surficial soil is a comprehensive process. In this study, the E312 chimney of the ILVA in Taranto was considered as the zero point (release point). After carefully fitting the data in Table 9 with the least square regression method, it was found that the relationships between the activity concentrations of 210 Po and 210 Pb in superficial soils and distance of the sampling sites from the release point can be well described by equations of polynomial ( Figure 2). From Figure 2, it was observed that (i) the activity concentrations of 210 Po and 210 Pb in the soils are increasing with the increase of the distance within 20 km and (ii) the activity concentrations can be considered as background level beyond 40 km. Therefore, from Figure 2, it was predicted that the maximum activity concentrations of 210 Po and 210 Pb in the territory could occur at about 20 km from the release point. These data can also provide useful information for the reasonable site selection and distribution for the future radiological survey. The annual precipitation rate in Puglia region is much less than other regions of Italy; therefore, dry deposition could play an important role for the long distance transportation of 210 Po and 210 Pb released from the ILVA in Taranto.

Polonium-210 and 210 Pb Concentrations in Atmospheric
Particulate. The atmospheric particulate samples in the first sampling campaign were taken with high volume samplers on 11 to 17 November 2008 at two sites located around the plant ILVA in Taranto, that is, Via Machiavelli and Cisi. At each site, two kinds of atmospheric particulate samples were collected, one with an atmospheric particulate mass concentration in the fraction of an aerodynamic diameter ≤10 m (PM 10 ), and another in the fraction of ≤2.5 m (PM 2.5 ).
As indicated in Table 10 At the first glance of the data in Table 10 and the weather record, the activity concentrations of 210 Po and 210 Pb were highly variable and similar to the mass concentration variation of the atmospheric particulate [7], in particular, depending on the variability of weather conditions encountered during the sampling period. In the raining days less particulates were collected, and, therefore, also lower activity concentrations of 210 Po and 210 Pb were detected.
In the UNSCEAR [12] reports that the reference concentrations in air are about 50 Bq m −3 (ranged from 12 to 80 Bq m −3 ) for 210 Po and 500 B m −3 (ranged from 28 to 2250 Bq m −3 ) for 210 Pb, respectively, and they are sites specific. It was reported that the yearly average concentrations of 210 Pb in surface air over Europe were about 200-700 Bq m −3 . Therefore, the obtained concentrations of 210 Po and 210 Pb in this study were in well agreement with the reported values.
Lead-210 and 210 Po in atmosphere come from several sources: (i) from volcanic dust [17,18], (ii) from 222 Rn gas which is exhaled from the ground into the atmosphere [19,20], (iii) from resuspended soil particles (see [21][22][23][24]), and (iv) from widespread dispersal of phosphate fertilizers, fossil fuel combustion, biomass burning, and industrial processes It was said that the concentrations of 210 Po in air are extremely high for weeks and in some cases months before major volcanic eruption [21,23]. Shortly after the eruption stops, over three orders of magnitude decrease in the activity of 210 Po in air even at >1300 km from the eruption site. Due to its volatility, 210 Po is highly enriched in volcanic gases with 210 Po/ 210 Pb activity ratios up to 600 (57 to 614) in the Stromboli's plume which has been reported [30]. As far as the case of ILVA in Taranto is concerned, the affection from the source of volcanic eruption was not considered, as there was no any volcanic eruption in the investigation period in the studied region and nearby. In fact, the 222 Rn gas exhalation (the second category) is the most important source contribution of 210 Po and 210 Pb in the obtained samples, which can be confirmed by the similar concentrations of 210 Po and 210 Pb and the 210 Po/ 210 Pb ratios in some samples if compared to that at the control site. As shown in Table 10, the activity concentrations of 210 Po and 210 Pb found in the fraction of PM 10 did not differ from the corresponding values found in the fractions of PM 2.5 ; thus, it was showed that almost all of 210 Po and 210 Pb were found only in the fraction of PM 2.5 . This conclusion is consistent with that reported by Martell and Moore [31], who observed that 90% of 210 Pb ion aerosols are associated with particles ≤0.3 m.
Studies showed that, when inhaled, smaller particles can be more toxic than a comparable mass of larger particles of the same material, as the health effects are directly linked to their bigger particle surface area and higher solubility [32]. Due to the fact that (i) the sizes of the resuspended soil particle in most cases are >2 m and the 210 Po/ 210 Pb ratios are high and in the range of 0.77-1.0 and (ii) the 210 Po/ 210 Pb ratios in the collected samples are much less than 0.77-1.0, therefore, it was speculated that the 210 Po and 210 Pb contribution from the resuspended soil particles (the third category) in the collected samples (Table 10) were negligible.
According to the literature [33], the atmospheric residence times of the fine aerosols are 33-66 days, and the calculated 210 Po/ 210 Pb activity ratios in air samples should be ≤ 0.092. In fact, from the sampling to measuring, 10-20 days are needed in the study, and taking into account the 210 Po decay and ingrowth from 210 Pb in that period, the 210 Po/ 210 Pb activity ratio of 0.122 obtained at the control site is reasonable. If 0.122 was taken as the reference value, it was seen that the 210 Po/ 210 Pb activity ratios in Table 10 were only a little bit higher than the reference value in nearly all the atmosphere particulate samples collected from the ILVA in Taranto As shown in Table 2, the mass concentration of particle in air samples collected in the ILVA in Taranto on 11-17 November 2008 ranged from 3.84 to 30.68 g m −3 with a mean value of 13.6±8.4 g m −3 . The obtained value was much lower than the annual mean values of 42 g m −3 at the site of Machiavelli and 34 g m −3 in Cisi obtained by the Monitoring Center of Taranto. The lower value was attributed to a number of factors: (i) during the sampling period, the wind direction was changed frequently and it was not possible to relocate quickly the equipment for collecting particulates in a short time; therefore, the equipment was not in the well position of downwind of the plant emission; (ii) during the sampling period, namely, on days 12, 13, and 14 November 2008, the plant was operated with only one production line of two and partially reduced the emission of particles; (iii) during the sampling period, two or three days of raining occurred and the process could surely affect the scavenge of the particulates emitted from the chimney of the plant and reduce the particulate concentration in air. As shown in Figure 4 Table 11, the obtained activity concentrations of 210 Po in samples of PM 2.5 and PM 10 were in the range of 154-564 Bq m −3 , those of 210 Pb in the range of 618-1087 Bq m −3 , and the 210 Po/ 210 Pb activity ratios in the range of 0.231-0.543 (mean: 0.343 ± 0.097). The maximum concentrations of 210 Po and 210 Pb were about 11.6 and 2.72 times as high as those at the control sites, respectively. It seems that a stable contamination factor for 210 Pb was found and that for 210 Po was much higher compared with the corresponding data in the first sampling campaign.

Estimation of the Committed Effective Dose for Members of the Public due to Inhalation of Air Contaminated with 210
Po and 210 Pb. The risk evaluation of 210 Po and 210 Pb for the general public could involve two aspects, that is, radiological and biological toxicities. In this paper only the radiological toxicity is evaluated. The possible radiation pathway could include the external exposure from cloud plume and ground and internal exposure due to inhalation of air and digestion of food and water contaminated with 210 Po and 210 Pb. Due to the fact that 210 Po is a pure -emitter and 210 Pb is a week -emitter, internal exposure through inhalation of the contaminated air in this case could be a very important pathway for individual of the public. For estimation of the annual committed effective dose ( , Sv y −1 ) for members of the public due to inhalation, the following equation was used: where is 210 Po or 210 Pb; is the mean or maximum activity concentration of 210 Po or 210 Pb in atmospheric particulate, Bq m −3 ; is the mean respiration rate for individual of different age group [34], m 3 d −1 ; is the exposure time, 365 d y −1 ; ,inh , the age-dependent dose coefficient for inhalation of particulate aerosols containing 210 Po or 210 Pb. ,inh also depends on the absorption rate of the materials inhaled. In general consideration, there could exist three types of 210 Po or 210 Pb materials in the taken atmospheric particulate samples, that is, F: fast absorption type, M: moderate type and S: slow type. In this study for sure to obtain the conservative dose results, the ,inh value for the material of slow type was selected for the dose evaluation [35].
Based on the mean or maximum concentrations of 210 Po and 210 Pb in the samples collected on 19-29 November 2008 in Castel Romano Roma and on 11-17 November 2008 as well as in 4-12 May 2009 at sites of the ILVA in Taranto, the estimated annual committed effective doses for different age group of the public were given in Tables 12-14.
Only taking into account the adult population, from Table 12 it was seen that (i) the mean committed effective doses due to intake of 210 Po and 210 Pb from the atmospheric particulate at the control site (Castel Romano Roma) were 1.69 and 18.1 Sv y −1 , respectively, and with a total dose of 19.8 Sv y −1 (ii) the dose from 210 Pb was 10.7 time as high as that from 210 Po, and (iii) the doses received for all age groups were nearly in the same level. As shown in Table 13, the mean committed effective doses for adult due to intake of 210 Po and 210 Pb from the atmospheric particulate at the sites of the ILVA in Taranto on 11-17 November 2008 were 3.62 and 28.3 Sv y −1 , respectively, and with a total dose of 31.9 Sv y −1 , and they were 1.6-2.1 times as high as that at the control site. Based on the maximum concentrations, the dose from 210 Po and 210 Pb could be 7.87 and 49.9 Sv y −1 , respectively, and with a total dose of 57.7 Sv y −1 . Therefore, about 12.1 Sv y −1 at mean and 37.9 Sv y −1 at maximum could be attributed to the 210 Po and 210 Pb emitted from the chimney of the ILVA in Taranto, of which about 84% were the contribution of 210 Pb. The estimated dose in this investigation campaign could be a underestimated value due to the particle scavenge effect of raining, the wind direction changing and the lower particle emitting rate during the plant production with only halfcapacity.
Based on the results in the second investigation campaign (Table 14), the mean committed effective doses for adult due to intake of 210 Po and 210 Pb from the atmospheric particulate at the sites of the ILVA in Taranto in 4-12 May 2009 were 10.1 and 36.8 Sv y −1 , respectively, and with a total dose of 47.0 Sv y −1 , and they were 2.0-6.0 times as high as that at the control site. If based on the maximum concentrations, the doses from 210 Po and 210 Pb could be 19.7 and 49.3 Sv y −1 , respectively, and with a total dose of 69.0 S y −1 . Therefore, about 27.2 Sv y −1 of the dose at mean and 49.2 Sv y −1 at maximum could be attributed to the 210 Po and 210 Pb emitted from the chimney of the ILVA in Taranto, of which about 63.5-69.0% was the contribution of 210 Pb.
Due to the involvement of man-made release of 210 Po and 210 Pb, the operational process of the ILVA in Taranto could be considered as a planned and prolonged exposure situation. The radiological risk evaluation to members of public in Taranto was based on the guidance from ICRP and the Italian law. The constrains and reference levels from ICRP to the public exposure from planned exposure situation is no more than 0.3 mSv y −1 from waste management operations [36] and is no more than 0.1 mSv y −1 from situation of prolonged exposure [37]. The guidance from ICRP is also the basic principle for the Italian law as well as its subsequent amendments [38]. The additional committed effective doses estimated to individuals of the public due to emission of 210 Po and 210 Pb from the plant of the ILVA in Taranto in the two times of investigation campaigns all were much lower than the established action level, even in the worst case (49.2 Sv y −1 ). Therefore, no specific interventions should be required, as far as only inhalation of 210 Po and 210 Pb was concerned.
It was reported that the average committed effective dose resulting from exposure to environmental background radiation in Italy, regardless of exposure to indoor radon, is greater than 1 mSv y −1 [39], and the global average human exposure from natural sources is 2.4 mSv y −1 [12,40]. From radiation protection point of view, it is predicted that the exposure damage to individuals of public due to the inhalation of 210 Po and 210 Pb released from the ILVA in Taranto could not be observable and significant.
Being time-, labor-, and money-consuming, the survey at this stage was mainly focused on studies of the radiological impact of 210 Po and 210 Pb through inhalation on the adult members of the public. According to literatures [5], 210 Po and 210 Pb are the most important radionuclides released from the coal power plant, steel-making industry, and refractory material industry, and inhalation is a very important exposure pathway. However, the other possible radiological impacts of uranium, thorium, or radium isotopes, on workers and the public in the region of the ILVA in Taranto resulting from the treatment, storage, disposal, and reuse of the produced waste through ingestion of the contaminated food, vegetable, water,   The respiration rate, annual respiration volume, and the 210 Po and 210 Pb dose coefficients used to calculate the committed effective dose for each age class are the same as those listed in Table 12. and soil and inhalation of air all can be subjects of future researches.
In spite of the difficulties to statistically observe the bioeffect of radiation dose at very low exposure rate, additional concentrations of 210 Po and 210 Pb from the man-made release are usually considered toxic to public due to their radiation damage and chemical toxicity. One of the three well-known radiation protection principles (justification, optimization, and dose limitation) recommended by the ICRP for practices is that doses to individuals of public and to occupationally exposed workers should be kept as low as reasonably achievable (optimization principle). Thus, for realizing the optimization principle for the practice, or-so-called "planned and prolonged exposure situation, " the administrators of the plant ILVA in Taranto should keep in mind to better control their production procedures, for instance, (i) selecting the raw materials containing lower activity of radionuclides, (ii) keeping the filtration system always in good function, (iii) inventing and testing new technologies to improve the effect of dust particle removal (e.g., wet method), and (iv) well disposal of the wastes to prevent the second contamination of the environment, mankind, and so forth.
Moreover, the biological and chemical toxicities and epidemiological survey of lead in the area of the ILVA in Taranto should also be a subject of study as far as the public health is concerned because the release of 17.8 t y −1 or even more of lead smoke into the atmosphere, estimated in this work, is not a quantity negligible.

Conclusion
The radiological survey on the Iron-and Steel-making Plant ILVA in Taranto was mainly focused on contamination source-term investigation and exposure impact evaluation of the volatile radionuclides 210 Po and 210 Pb. The activity concentrations of 210 Po and 210 Pb in the raw materials, dust particles, surficial soils, and atmospheric particulate samples collected in the area of ILVA in Taranto were determined. The results showed that the activity concentrations in the raw materials were in the range of 3.46-17. Pb in the atmospheric particulate are one to three order of magnitude higher than that found in the top soil. In the worst case detected, the mean committed effective doses for adult due to intake of 210 Po and 210 Pb from the inhaled atmospheric particulate at the sites of the ILVA in Taranto were about 10.1 and 36.8 Sv y −1 , respectively, and with a total dose of 47.0 Sv y −1 , and they were 2.0-6.0 times as high as that at the control site. Based on the maximum concentrations, the doses from 210 Po and 210 Pb could be 19.7 and 49.3 Sv y −1 , respectively, and with a total dose of 69.0 Sv y −1 . After deduction of the background contribution, about 27.2 Sv y −1 of the dose at mean and 49.2 Sv y −1 at maximum could be attributed to the 210 Po and 210 Pb emitted from the E312 chimney of the ILVA in Taranto, of which about 63.5-69.0% was the contribution of 210 Pb. The constrains and reference levels from ICRP to the public exposure from planned exposure situation is no more than 0.3 mSv y −1 from waste management operations and is no more than 0.1 mSv y −1 from situation of prolonged exposure. Therefore, no specific interventions should be required, as far as only inhalation of 210 Po and 210 Pb was concerned.