- Email: [email protected]
The potential radiological impact from a Brazilian phosphate facility
Journal of Environmental Radioactivity 136 (2014) 188e194
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
The potential radiological impact from a Brazilian phosphate facility cio Glo ria dos Reis*, Dejanira da Costa Lauria Ro ~o e Dosimetria (IRD/CNEN), Avenida Salvador Allende, s/n, 22780-160 Rio de Janeiro, RJ, Brazil Instituto de Radioproteça
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 August 2013 Received in revised form 6 May 2014 Accepted 3 June 2014 Available online 24 June 2014
In the semiarid region of Brazil, a facility for the production of phosphoric acid for fertilizer is in the last stages of the planning phase. The raw feedstock of Santa Quiteria has a very high level of uranium associated with the phosphate in form of apatite. The reaction by which phosphoric acid is produced generates phosphogypsum (PG) as a by-product. The ratio of phosphogypsum to phosphoric acid is approximately 5 to 1. After all of the phosphate has been extracted and processed, it is expected that some 37 million tons of phosphogypsum will be produced, containing 13 Bq/g of 226Ra and 11 Bq/g of 210 Pb. To assess the potential impact of this PG stack on the surrounding inhabitants, a generic impact assessment was performed using a modeling approach. We estimated the amount and shape of the residue stack and used computational codes for assessing the radiological impact in a prospective risk assessment. A hypothetical farmer scenario was used to calculate two potential doses, one near the site boundary and another directly over the stack piles after the project is shut down. Using a conservative approach, the potential public dose was estimated to be 2.8 mSv/y. This study identified the rainfall erosion index, dissolution rate of PG, radionuclide distribution coefficients and fish consumption rate as parameters where improved information could enhance the quality of the dose assessment. The disposal and shape of the stack is of major concern, since the PG erosion might be the main pathway for the environmental contamination; therefore, studies should be carried out to determine a suitable shape and disposal of the stack. Furthermore, containment barriers should be evaluated for their potential to reduce or avoid environmental contamination by runoff. In addition, the onsite public dose underscores the importance of a planning for remediation of the area after the plant is shut down to assure that neither the public nor the environmental health will be affected by the presence of the PG stack. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Phosphogypsum Radiological impact Modeling NORM
1. Introduction Impact assessment is a process of identifying the consequences of a past, planned or current activity. This process has been divided in two types: retrospective and prospective assessments. A retrospective assessment quantifies the impact of an ongoing activity or tries to understand the impact of a past activity, while a prospective assessment should be developed during the planning phase of a project. The overarching objective of an impact assessment is to derive a mitigation plan to minimize or eliminate negative impacts while maximizing positive impacts. It is a requirement for licensing a new facility in many countries. For the nuclear industry, a radiological impact assessment is required in addition to the conventional impact assessment and should be carried out for all phases of a project. This assessment
* Corresponding author. Tel.: þ55 21 2173 2804; fax: þ55 21 2173 2701. ria dos Reis), E-mail addresses: [email protected], [email protected] (R. Glo [email protected] (D. da Costa Lauria). http://dx.doi.org/10.1016/j.jenvrad.2014.06.001 0265-931X/© 2014 Elsevier Ltd. All rights reserved.
estimates the consequences to humans of releasing radioactive isotopes to the environment. The process is multidisciplinary and requires several areas of expertise to identify and quantify the source term, to predict the environmental transport and to estimate the human exposure to radiation and the resulting total dose, which comprises both the internal and external dose (Till and Grogan, 2008). The estimated human dose value is then compared with the adopted value of the dose limit. The level of radionuclide discharge to environment is controlled so that the dose limit is not exceeded. Any manipulation of materials that exceed the dose limit need to be controlled by a regulatory body. Apart from the nuclear industry, only industries that generate Naturally Occurring Radioactive Material (NORM) have to perform radiological impact assessments. The need to perform these assessments is a challenge for NORM industries, which must comply with standards from nuclear industries in addition to the usual environmental and worker safety requirements. A list of industry sectors have been identified as the most likely to produce NORM: oil and gas production, the niobium and ironeniobium industry, manufacture of titanium dioxide pigments, the
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
phosphate industry, the zircon and zirconia industry, production of tin, copper, aluminum, zinc, lead, and iron and steel, combustion of coal and water treatment (IAEA, 2006). The risk level for the public and the environment that can be attributed to these industries is not yet well known, not only because they differ from each other but also because the quantity of radionuclides released to the environment or contained in process residues depends on the original levels in the feedstock minerals and on the process used by the industry (UNSCEAR, 2008, IAEA, 2012). Among the NORM industries, the phosphate fertilizer industry plays an important role in conventional and radioactive pollution. Phosphate rock is used worldwide for manufacturing phosphoric acid and chemical fertilizers. It is well known that this type of rock may contain elevated levels of uranium and its decay products, while its thorium and potassium concentrations are comparable to those usually found in soils. Mortvedt (1994) reported uranium concentrations in phosphate rock worldwide ranging from 3 to 400 mg/kg, or 37e4900 Bq/kg. For 226Ra, a range from 100 to 10,000 Bq/kg has been reported (Rossler et al., 1979). The chemical process for producing phosphoric acid redistributes the radioactive isotopes among the products and residues in such way that around 90% of the U will usually be found in the phosphoric acid fraction. In contrast, 90e80% of the radium isotopes 210Pb and 210Po will be transferred from rock phosphate raw material into the phosphogypsum by-product (Rutherford et al., 1995; Hul and Burnet, 1996; Mazzili et al., 2000). Mazzili et al. (2000) reported that levels in the phosphogypsum varied between 22 and 695 Bq/kg for 226Ra, between 47 and 894 Bq/kg for 210Pb, between 53 and 677 Bq/kg for 210 Po and between 7 and 175 Bq/kg for 232Th. Therefore, radioactive isotope content of the products and residues of fertilizer production can affect not only the worker health but also the environment. The International Agency of Energy Atomic (IAEA) suggests that control and regulation of activities leading to exposure to NORM is not necessary if the effective radiation dose received by a worker or member of the public does not exceed approximately 1 mSv in a year. The IAEA (2005b) also establishes the activity concentration levels in specific materials that do not require any regulation of the related activities (1 Bq/g for uranium and thorium series radionuclides and 10 Bq/g for 40K) (IAEA, 2011). In Brazil a standard for regulating the mining and milling industries establishes a criterion for classification of the NORM industry according to reference levels of concentration varying between 10 and 500 Bq/g and a dose increment of 1 mSv/y over natural background exposure. According to this standard, the severity of regulatory control for waste management and environmental protection increases with the increase of the radionuclide concentrations in material and the resulting doses. At the moment this standard is under review (CNEN, 2005). This research supports a radiological impact assessment of the Santa Quiteria fertilizer industry (originally known as Itataia mine), which is still in its planning phase. Using a computational code that was developed for assessing the environment impact of contaminated sites, this study aims to estimate the concentrations of isotope radionuclides in various environmental media, identify the main pathway responsible for the contamination and discuss actions to minimize the potential contamination.
Based on exposure scenarios, the exposure pathways are identified. Once scenarios are defined and exposure pathways identified, a basic conceptual understanding of the system is developed and a method to estimate the dose should be chosen. For the assessment of dose, suitable environmental parameters should be identified and selected, based on site specific data whenever possible. To reduce uncertainty about the pathways and parameters, a sensitivity analysis should be performed, which may lead to acquiring more data about the site. 2.1. Source term characterization e overview of Santa Quit eria's process The Santa Quiteria raw feedstock has a very high level of uranium associated with phosphate in the form of apatite. The deposit is estimated to have 80 million tons of ore resources, with 8.9 million tons of phosphate and 62.9 thousand tons of uranium oxide s, 1984; Alc^ (IPT, 1981; Nuclebra antara e Silva and Rosa, 1988; Fukuma, 1999). According to Saad (1995) and Saad and Chamon ria project will mine 1.7 million metric (2002), the Santa Quite tons of raw material that will produce 1.5 million tons/year of phosphate rock. ria's beneficiation consists of The procedure of Santa Quite breaking up the ore by drilling and blasting, crushing the rock, and then transporting the milled ore to the beneficiation plant to separate sand and clay and to remove impurities (Fig. 1). In the plant, the material is treated by a process called desliming, where hydrocyclone equipment is used to separate the fluid mixture into two components of differing densities. From this step, an estimated 1.4 million tons of finer particles containing 10.2 Bq/g of 238U will be generated (Fukuma, 1999). Afterwards, the phosphate matrix is further upgraded by a flotation process that uses a
2. Methodology In a radiological environmental impact assessment, the first step involves gathering and evaluating existing data and information about the site, including the source term, which should be identified and characterized both qualitatively and quantitatively. Next, taking into consideration the land use, diet and habits of the inhabitants of the region, exposure scenarios should be identified.
189
Fig. 1. Flow chart of the process.
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
190
turbulent water system, chemicals and air bubbles to float the phosphate particles to the water surface, where they are skimmed off. Reis Junior et al. (1987) have estimated the content of U3O8 in this waste as 297 mg/kg, which corresponds to a 238U concentration of 3 Bq 238U/g of material. Following these steps, the beneficiated rock is transferred directly to the chemical processing plant. The production of phosria will be carried out by the wet process phoric acid in Santa Quite method that includes digestion, filtration and concentration. Sulfuric acid is used for the feedstock decomposition. By this method, the beneficiated phosphate rock is dissolved in sulfuric acid to chemically digest the calcium phosphate. This step produces a slurry that is composed of phosphoric acid solution and a suspended solid, calcium sulfate, commonly known as phosphogypsum (PG). The slurry from this digestion is filtered and the solid phase is collected, washed, and sent to phosphogypsum stacks. The phosphoric acid produced will contain most of the uranium and a separation process will be used to split uranium from the acid phase. According to Fukuma (1999), for each ton of phosphoric acid ria, approximately 4.2 tons of phosphoproduced in Santa Quite gypsum will also be produced. Accordingly, if all 8.9 million tons of P2O5 is processed, then 37 million tons of phosphogypsum will be produced. The 226Ra and 210Pb contents of this phosphogypsum are estimated to be 13 Bq/g and 11 Bq/g, respectively (Fukuma, 1999). Therefore, three main processed materials will be produced and stored at the SQ site: a) 1.4 million tons of clay that contain 10.2 Bq 238U/g; b) The flotation tailings with a 238U level of 3 Bq/g; c) 37 million tons of phosphogypsum containing 13 Bq/g of 226 Ra and 11 Bq/g of 210Pb. For this research, we selected PG as the source term because of its quantity and high radionuclide concentration. 2.2. Study area ria site is located in the northeastern region of The Santa Quite Brazil, 4 340 2700 S, 39 470 1900 W, approximately 210 km from Fortaleza city. It consists of a 40.4 km2 area that is 6250 km long and 10,000 km wide (Fig. 2).
The Semiarid climate is warm with a mean annual temperature of 27 C (varying between 21 and 33 C). Rainfall is scarce but often of high intensity; more rainfall occurs between January and April, and there is a regular drought between May and December, in which the rainfall is frequently zero. The literature reports annual mean precipitation of approximately 824 mm for the period between 1984 and 1988. Most rainfall occurs in a very short period within one month. Therefore, runoff is highly seasonal. For the period between 1984 and 1988, the reported real mean of annual gross evapotranspiration was approximately 653 mm (Lopes Filho, 1977; Silva, 2004). The site topography is characterized by surface relief and high surface elevations. The Sertaneja depression and the Residual plateaus are the two major morphological units. The ore deposit is a recessed compartment embedded between the plateaus. It has an altitude that ranges between 350 and 580 m, whereas the residual u hill is plateaus have altitudes ranging from 650 to 1000 m. The Ce the highest peak in the region with an altitude of 1085 m. This hill forms the watershed between the Curu river and Groaíras river catchments (Silva, 2004). The small rivers that drain the site are not perennial, and rivers sustain flow only during rainy season. A dam named Quixaba with a total storage capacity of 2,300,000 cubic meters is located in the site. The vegetation reflects the differences in the altitude levels. On the high altitude areas, the vegetation is remnant dry forest, and the typical vegetation of the low altitude area is savanna (Gonçalves Junior and Souza, 2012). The soils of the area are mainly from the amendment of granites and gneisses “in situ” and the colluvial materials mobilized from these rocks (CPRM, 2008). The soil is predominantly an association of Red-Yellow, Alfisols and Entisols with a low permeability (Gonçalves Junior and Souza, 2012). The common feature of the various soils are their thinness, almost always less than 1 m, that the surface horizons are uniformly sandy and that the sotopostos horizons are all sandy-clay or clay (CPRM, 2008). The groundwater flows from the south to north-northeast (Saad, 1995). The water level stays between 5 and 60 m, averaging 30 m throughout the area. The recharge of the area depends entirely on rainfall that flows on the surface and subsurface. Two main factors influence the flux: the system of open sub-vertical fractures and the topography (Silva, 2004).
ria Site. Fig. 2. Location of the PG stack and the Critical Group in the Santa Quite
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
The area surrounding the site is sparsely populated and relaria, 45 km away. The tively remote; the nearest town is Santa Quite population's economic activities are limited to scattered subsistence farms that are used for limited agricultural activities (e.g., cultivation of beans, sugar cane, corn and manioc). Much of the area is used to graze cattle and sheep. The groundwater in the area is used to supply drinking water and small scale irrigation for agriculture, whereas the surface water supports cattle watering. 2.3. Definition of exposure scenarios An exposure scenario consists of a set of human activity patterns that are related to a range of exposure pathways and include an assumed duration for each pathway. The scenarios should encompass all typical exposure situations. Therefore, each scenario represents a range of exposure situations and encompasses a set of parameter values. Through models that represent the behavior and transfer of radioactive isotopes in the environment, these scenarios relate the activity concentrations in the material to individual doses (IAEA, 2005a). The farmer scenario is the most conservative scenario, in which the land use is intensive and all exposure pathways are considered. Since this scenario attempts to consider all possible exposure pathways, this simulation produces the highest values for radiation dose. In the farmer scenario, people might be assumed to live and work on the site, grow their own food on the site, and eat fish and drink water from the local water body. ria, we developed two scenarios for the For Santa Quite simulation: i) A farmer scenario where people live close to the borders of the site (Fig. 2). According to the land use, wells supply their drinking water and the crops and animal products are locally produced. The water supply for cattle is from the dam, and the irrigation water is from wells. The worst case scenario was simulated, whereby the groundwater flows from the PG stack toward the farm and the preferential direction of the wind is also from the stack toward the farms. The exposure pathways considered are direct external gamma radiation, radon and dust inhalation, drinking water and vegetable, milk, meat, aquatic food and soil ingestion. ii) Occupation of the site by the local inhabitants after the closure of the site and subsequent abandonment. The local people are living on the top of the PG stack. This is also a farm scenario, where the exposure pathways are external gamma, radon and dust inhalation, drinking water and plant, milk, meat, aquatic food and soil ingestion.
2.4. The model for estimating the dose-RESRAD codes The chosen model for estimating dose was the pathway analysis by RESRAD codes. RESRAD is a multiple pathway analysis computer code, comprising a set of mathematic models that attempt to estimate the radiation dose and risk for an individual located either onsite or outside the contaminated zone. The contaminated zone can encompass both surface and near surface buried waste. The RESRAD codes were designed for evaluation of radiological doses to a receptor from exposure to RESidual RADioactive materials in soil (Yu et al., 1993, 2001, 2003, 2006). The RESRAD 6.5 code (onsite) is able to estimate the radiation dose of an individual who spends time directly on the primary contamination (onsite), could consume crops and animal products that are grown on the contaminated site, and drinks water from well or/and surface water located on the contaminated site. On the other hand, RESRAD offsite is able to estimate the dose of an individual located away from
191
Table 1 Parameters used for the assessments. Parameter
Value
Reference
Hydraulic conductivity for saturated soil Hydraulic conductivity for unsaturated soil Hydraulic conductivity for phosphogypsum Dry bulk density of PG Unsaturated zone thickness Gross evapotranspiration Mean annual rainfall
24.2 cm/h ¼ 2120 m/y
Ferreira et al., 1999
1.6 103 m/h (13.9 m/y) 220 cm/d ¼ 803 m/y
Local data, (Silva, 1997, 2004) Van Alphen et al. (1971)
1.15 g/cm3 30 m
Van Alphen et al. (1971) Local data, (Silva, 1997, 2004) Local data, (Silva, 1997, 2004) Local data, (Silva, 1997, 2004) Fukuma, 1999.
Radionuclide concentrations in PG
Amount of PG in the stack Shape of stack
653 mm/y 824 mm 210
Pb ¼ 210Po ¼ 11 Bq/g; Ra ¼ 13 Bq/g; 228 Ra ¼ 0.41 Bq/g 228 Th ¼ 232Th ¼ 0.026 Bq/g; 230 Th ¼ 1 Bq/g; 234 U ¼ 238U ¼ 0.4 Bq/g 37 million tons 226
3400 m 500 m wide 50 m high
Saturated zone thickness Wind velocity
200 m 2.46 m/s
PG porosity Temperature
0.7 27 C
Fukuma, 1999; INB, 2013 Estimated based on Kedainiai's stark (Kardasecicius, 2006) and Kleinschmidt, 1992. Local data, (Silva, 1997, 2004) Estimated based on Amarante et al. (2001) ~ as et al. (2007) Duen Local data, (Silva, 1997, 2004)
the primary contamination (offsite). This individual could consume vegetables and animal products from agricultural fields that are contaminated by material from the primary contamination. Likewise, his/her drinking water can be from contaminated well or surface water body located away of the contaminated site. In addition to the radiation dose and risk, the radionuclide concentration in the environmental media and soil cleanup guidelines are also determined by the codes. For the assessments, the codes consider all the major exposure pathways by which an individual may be exposed: direct external radiation; inhalation of dust and radon; and ingestion of vegetables, meat, milk, aquatic food, and soil. To estimate the contamination of offsite locations, the code models the transport from the primary contamination to agricultural areas, pastures, a dwelling area, a well, and a surface water body. The main movements of the contaminants as considered by the code are the following: i) Leaching of radionuclide caused by water that flows down through the primary contamination and causes the contamination of the groundwater. The levels of radionuclides in groundwater and their dispersion are estimated by a rate-controlled transport and a three-dimensional dispersion groundwater flow models; ii) The atmospheric dispersion of dust as modeled by equilibrium models (Plume-rise and Gaussian plume models), and iii) erosion of contaminated material by surface runoff, which is modeled as a release to the surface water body. RESRAD is able to perform sensitivity and probabilistic analysis to evaluate the influence of input parameters and generate plots. Age-specific dose factors are included in the code's database. A set of more than ninety hydrogeological, meteorological, geochemical, geometrical (size, area, depth), and material-related (soil, concrete) default parameter values are provided in the RESRAD code (Yu et al., 1993a,b). According to the authors, these
192
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
default values have been carefully selected and are realistic, although conservative, and should be used for generic modeling. For more accurate use of the code, site-specific data should be used whenever possible. The codes are user friendly, available free of charge and can be downloaded from the RESRAD Web site (http://www.evs.anl.gov/ resrad). 2.5. Parameters and input data For the methodology herein adopted, although very important the choice of parameters is not a determining factor, since the use of sensitivity analysis should point those key parameters that have most influence on the dose and should be better studied or determined. A survey of the literature was performed to set local parameters. Hence, some of the RESRAD default parameter values were changed according to the available local parameters or to parameters that best described the local characteristics (Table 1). Consequently, hydraulic conductivity, density and porosity values for PG from the literature were introduced to reflect PG characteristics. The estimated shape of the stack of 37,000,000 tons of PG was based on the PG stack in Kedainiai, which encompasses an area of 84 ha and is 60 m high, and holds 22,000,000 tons of PG (Van ~ as et al., 2007; Kardasecicius, 2006). The Alphen et al., 1971; Duen concentrations of the radionuclide daughter of short half-life were assumed to be in secular equilibrium with its parent. 3. Results and discussion 3.1. Assessment of the dose for the farm scenario located closest to the site Having taken into account all exposure pathways, the simulation using the RESRAD Offsite code for the farm located close to the site reaches a dose value for the critical group dose of about 2.5 mSv/y at 10 years and rising 2.8 mSv/y at 70 years (Fig. 3). The main contributor to this dose is 226Ra, which is responsible for 91% of the dose, followed by 210Pb. The other radionuclides did not contribute significantly to the dose (Fig. 3, Table 2). Taking into consideration the transport of contamination by environmental media, the simulation found the water-dependent pathways to be responsible for most of the dose (92%), and among those, release from the source to surface water is the predominant contamination pathway. No significant transport of contamination through groundwater was predicted by the model. The atmospheric pathway contributes 7% to the total dose, mainly due to the release of radon from the PG stack, followed by its
Fig. 3. Contribution of the radionuclides for the total dose, considering all exposure pathways.
Table 2 Total Contribution for individual radionuclides, t ¼ 70 years. Radionuclide
Total dose (mSv/y)
210
0.25 (9%) NR 2.45 (90%) 2.44 E06 1.89 E15 5.77 E03 3.97 E03 1.2 E04 1.18 E04
Pb 210 Po 226 Ra 228 Ra 228 Th 230 Th 232 Th 234 U 238 U NR ¼ not reported.
inhalation. Therefore, dust dispersion by wind was not signaled as an important contamination pathway. Among the exposure pathways, consumption of fish made the largest contribution to the total dose (86%), followed by radon inhalation, which contributed 7%. All together, the other pathways were responsible for less than 7% of the total dose (Fig. 4 and Table 2). Sensitivity analysis is a technique used to identify sensitive parameters, e.g., those having the greatest impact on dose assessment results, and to explore potential reductions in the uncertainties associated with dose parameters. The sensitive analysis, which was carried out for all parameters, noted that the rainfall erosion index was the main parameter influencing the dose. A twofold variation in the parameter value doubled or tripled the dose value (Fig. 5). Hence, according to the model, the erosion caused by rainfall might be responsible for the contamination of the surface water and consequently for the fish contamination. Based on this result, the rainfall erosion index is a parameter that should be set based on reliable site specific information data. If surface water contamination is predicted to be the major secondary source of contamination, some parameters related to the PG material dissolution in water and the radionuclide sorption by sediments might also be determined experimentally. Considering the rainfall regime in the SQ region (intense rain in a short period of time), the low permeability of soil and the scarce vegetation of the site, the erosion of material and its displacement into surface water might be a real possibility. Also the dissolution of the material in body water is a real possibility, in view of the solubility coefficients of different forms of calcium sulfate in phosphogypsum determined by Santos, who found that 2.8 g of phosphogypsum would be completely solubilized in 2000 L of water (Santos, 2002). The contamination of surface water might lead to contamination of sediment. Although sediment is not a pathway for human in the herein considered scenarios, the scenarios may change in the
Fig. 4. Contribution of the exposure pathway for the total dose.
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
Fig. 5. Influence of the erosion index caused by rainfall on the dose value.
future and sediment might be used for agricultural and building purposes for example (Rochedo et al., 2002). Also the contamination of water body may affect the local biota and an assessment of biota dose should be performed. Based on the results, the disposal and shape of the PG stack is of major concern because runoff might be the main pathway for surface water contamination. 3.2. Assessment of dose for the occupancy scenario after the closure and abandonment of the site For the scenario that considers occupancy of the site after the closure of the facility, the total dose reached the value of 160 mSv/ year at 70 years after the closure of the site, Fig. 6. In this case, the main contributor for the dose is 226Ra (222Rn), followed by 210Pb. Inhalation of radon contributes most of the total dose. Our results lead to clear conclusions. According to the current recommendations by IAEA and to Brazilian standards, the activity concentrations of the PG produced will exceed the nuclide specific exemption values of 1 Bq/g. The dose criterion for the public of 1 mSv/y is also surpassed. For that reason, the use of this material is not unrestricted, and the material would not be exempt from regulatory control. Therefore, it is unlikely that the PG produced could be used as amendment in agriculture or for construction. After the closure of this facility, the legacy waste will require strict control of the material into the foreseeable future. The total costs related to the future management should be considered in the feasibility study for the facility. Therefore, alternatives for the use and production of the PG should be considered to minimize the impact and cost of waste management. 4. Conclusion In the exposure scenarios, we considered PG as the only source term of radiation, and even so, the critical group dose was found to
193
exceed the public dose limit of 1 mSv/year (as established by the CNEN NN-3.01 regulation known as the Brazilian Basic Radiation Protection Guidelines, which is based on IAEA standards). Based on these results, the disposal and shape of the stack is of major concern because the erosion caused by rainfall may be the primary pathway for environmental contamination. Therefore, studies should be carried out to determine the parameters related to material erosion and the shape and disposal of the stack. Additionally, containment barriers should be evaluated for their potential to reduce or avoid environmental contamination by runoff. The second, or post-operational, scenario reveals a high dose for inhabitants who settle on the PG stack in the case of site abandonment. This underscores the need to plan the decommissioning of the site from the planning phase, to include this subject in the feasibility study for the project and to examine alternatives for the safe disposal or recycling of PG. In summary, the high doses under all scenarios in this study emphasize the importance of conducting an environmental risk assessment before a facility begins to operate. Because the area may eventually be reused after the closure of the facility, it is also necessary to plan for waste management to prevent potential liabilities. As IAEA has been be emphasizing the economic and social costs of remediation are incomparably higher than prevention, the results of this research point out the needs for the owner of the industry avoid possible future costs that can be unsustainable. Planning for the decommissioning of the area must occur in the initial design phase, before the construction of the facility, because measures to minimize risk and liability must be evaluated and their related costs included. It is crucial to plan for all possible future liabilities when considering the future decommissioning of the ria facility. Santa Quite The data generated and presented in this simulation are based on a conservative approach. For example, some parameters were the RESRAD defaults, and their uncertainties were not considered. The simulation noted the importance of the water pathway for environmental contamination. In light of its importance and to decrease the uncertainty in the results, site specific parameters related to this pathway should be determined whenever possible. Nevertheless, this methodology is useful for estimating risks to the public and workers in the decision-making process related to the planning and sustainability of the project. This work shows the feasibility of using a code that is designed to assess the impact of residual contamination proactively to assess the impact assessment of a facility in the planning phase of the project. Acknowledgements The authors would like to thank the CNPq (National Research Council) Brazilian funding agency for its financial support. The authors gratefully acknowledge the IAEA for promoting the improvement of skills in the field of assessment of environmental radiological impact through its EMRAS and MODARIA programmes, in which the authors are involved. Participation in these programmes has been of great importance for accomplishing this work. References
Fig. 6. Contribution for the total dose of the different exposure pathways.
^ntara e Silva, J.R., Rosa, C.V., 1988, Caracterizaça ~o geote cnica, geolo gica, química Alca rio te cnico específico. Nuclebra s, 131p. e reservas da jazida de Itataia: relato lico Brasileiro, Amarante, O.A.C., Brower, M., Zack, J., 2001. Atlas do Potencial Eo Ministerio das Minas e Energia. Brasilia, Brasil. http://www.cresesb.cepel.br/ publicacoes/download/atlas_eolico/Atlas%20do%20Potencial%20Eolico% 20Brasileiro.pdf. ~o Nacional de Energia Nuclear, 2005. CNEN-NN-4.01, Requisitos de SegurComissa ~o Radiolo gica para Instalaço ~es Minero-Industriais, Brasil. http:// ança e Proteça www.cnen.gov.br/seguranca/normas/pdf/Nrm401.pdf.
194
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
sicas de Comiss~ ao Nacional de Energia Nuclear, 2011. CNEN-NN-3.01, Diretrizes Ba ~o Radiolo gica, Brasil. http://www.cnen.gov.br/seguranca/normas/pdf/ Proteça Nrm301.pdf. rio de Minas e Energia, Secretaria CPRM, 2008. Programa Geologia do Brasil. Ministe ~o Mineral CPRM e Serviço Geolo gico do de Geologia, Mineraç~ ao e Transformaça Brasil. Contrato CPRM- UFC N . 058/PR/05. ~ as, C., Liger, E., Can ~ ete, S., Pe rez, M., Bolívar, J.P., 2007. Exhalation of (222)Rn Duen from phosphogypsum piles located at the Southwest of Spain. J. Environ. Radioact. 95 (2e3), 63e74. ^ncia da Mineralogia da Fraça ~o Ferreira, M.M., Fernandes, B., Curi, N., 1999. Influe ~o Sudeste do Brasil. Rev. Argila nas Propriedades Físicas de Latossolos da Regia ^ncia Solo 23, 515e524. Brasil. Cie ~o de Radionuclideos Naturais nos Produtos Obtidos Fukuma, H.T., 1999. Distribuiça ~o de Acido rico a Partir da Rocha Fosfa tica de Itataia e Cear na Fabricaça Fosfo a. MSc. Thesis. Universidade de S~ ao Paulo, Brazil. ~o ambiental de Santa Gonçalves Júnior, J.T., Souza, M.J.N., 2012. Caracterizaça ria, Ceara : a nova cidade Uranífera do Brasil. Rev. Geonorte Ediça ~o Especial Quite 2 (4), 1368e1377. Hul, C.D., Burnet, W.C., 1996. Radiochemistry of Florida phosphogypsum. J. Environ. Radioact. 32, 213e238. IAEA, 2012. Environmental modelling for radiation safety (EMRAS). A summary report of the results of the EMRAS programme (2003e2007). IAEA-TECDOC1678. http://www-pub.iaea.org/MTCD/publications/PDF/TE_1678_Web.pdf. IAEA, 2005a. Guidebook on Environmental Impact Assessment for In Situ each Mining Projects. IAEA-TECDOC-1428, p. 163. IAEA, 2005b, Derivation of activity concentration values for exclusion, exemption and clearance, safety reports series No. 44. IAEA, 2006. Assessing the Need for Radiation Protection Measures in Work Involving Minerals and Raw Materials, Safety Series, vol. 49. IAEA, Vienna, p. 55. IAEA, 2011. International basic safety standards for protection against ionizing radiation and for the safety of radiation sources. Interim edition. http://wwwpub.iaea.org/MTCD/publications/PDF/p1531interim_web.pdf. Industrias Nucleares do Brasil e INB, 2013. http://www.inb.gov.br/pt-br/WebForms/ interna.aspx?secao_id¼98. gicas do Estado de Sa ~o Paulo (IPT), 1981. Relato rio n. Instituto de Pesquisas Tecnolo gico-geote cnicos para projetos de engenharia na Mina de 15-253. Estudos geolo ria/CE, vol. 1. INB, S~ Itataia: município de Santa Quite ao Paulo. Kardasecicius, L., 2006. Treakearth. http://www.trekearth.com/gallery/Europe/ Lithuania/South/Kauno/Kedainiai/photo433937.htm. Kleinschmidt, D., 1992. Phosphogypsum stack closure. In: Proceedings of the Workshop Phosphate Fertilizers and the Environment, pp. 185e192. USA. gicos para a regia ~o de Itatira. Lopes Filho, A.F., 1977. Bases de estudos hidrogeolo cnica. Nuclebra s/DRM, Fortaleza. Nota te Mazzili, B., Palmiro, V., Saueia, C., Nisti, M.B., 2000. Radiochemical characterization of Brazilian phosphogypsum. J. Environ. Radioact. 49, 113e122. Mortvedt, J.J., 1994. Plant and soil relationships of uranium and thorium decay series radio e nuclides e a review. J. Environ. Qual. 23, 643e650. 1984. Jazida de Itataia: relato rio de pesquisa mineral. INB, Fortaleza, NUCLEBRAS, p. 330.
rio Reis Junior, J.B., Aquino, J.A., Luz, I.L.O., 1987. Estudo em Planta Piloto do Mine sforo-uranífero de Itataia e CE. NUCLEBRAS/CDTN. Fo Rochedo, E.R.R., Amaral, E.C.S., Hacon, S.S., 2002. Uncertainties in the definition of critical groups to chronic exposure scenarios. In: 5th International Conference on High Levels of Natural Radiation and Radon Areas: Radiation Dose and Health Effects, Munich. Excerpta Mediaca e International Congress Proceeding Series, vol. 1225. Elsevier Science B.V, Amsterdam, pp. 25e31, 2000. Rossler, C.E., Smith, Z.A., Bolch, W.E., Prince, R.J., 1979. Uranium and radium-226 in Florida phosphate materials. J. Health Phys. 37, 269e277. Rutherford, P.M., Dudas, M.J., Arocena, J.M., 1995. Radioactivity and elemental composition of phosphogypsum produced from three phosphate rock sources. Waste Manag. Res. 13, 407e423. gico na Indústria de Fertilizantes Saad, S., 1995. Alguns Aspectos do Impacto Radiolo Fosfatados e Estudo de um Caso-Itataia (Ce). MSc. Thesis. Instituto de Engenharia Militar, Brasil. todo eletromagne tico Saad, S., Chamon, N., 2002. Levantamento geofísico com o me : (VLF) e eletrorresistividade na area da jazida de Itataia do Estado do Ceara rio de prospecça ~o geofísica. Texto, anexos e Mapas. ENGERAD, Rio de relato Janeiro, p. 86. gico ambiental do fosfogesso Santos, A.J.G., 2002, Avaliaç~ ao do impacto radiolo brasileiro e lixiviaç~ ao de 226Ra e 210Pb. 2002. 165 f. Tese (Doutorado) e ticas e Nucleares, Sa ~o Paulo, Brasil. Instituto de Pesquisas Energe ^mico- Ambientais do Uso do Fosfogesso na Silva, L.H. da C., 1997. Aspectos Econo Agricultura. MSc. thesis. Universidade Federal do Rio de Janeiro, Brasil. ~o Hidrogeolo gica da Jazida de Itataia-CE, XIII Silva J.R.A.E., 2004, Caracterizaça ^neas e Cuiab Congresso Brasileiro de Aguas Subterra a, Brasil. Till, J.E., Grogan, H., 2008. Radiological Risk Assessment and Environmental Analysis. Book. Oxford University Press Inc, New York, USA. United Nations Scientific Committee on the Effects of Atomic Radiation e UNSCEAR, 2008. Report to the General. Assembly, with scientific annexes. Vol. I: Report to the General Assembly, Scientific Annexes A and B. Van Alphen, J.G., De Los Rios Romero, F., 1971, Gypsiferous Soils: Notes on Their Characteristics and Management, International Institute for Land Reclamation and Improvement, Bulletin 12, Netherlands. Yu, C., Loureiro, C., Cheng, J.-J., Jones, L.G., Wang, Y.Y., Chia, Y.P., Faillace, E., 1993a. Data Collection Handbook to Support Modeling Impacts of Radioactive Material in Soil. Environmental Assessment and Information Sciences Division, Argonne Laboratory, Argonne, Illinois, USA. Yu, C., et al., 1993b. Manual for Implementing Residual Radioactive Material Guidelines using RESRAD. Version 5.0, ANL/EAD/LD-2. Argonne National Laboratory, Argonne, Illinois, USA. Yu, C., et al., 2001. User's Manual for RESRAD Version 6. ANL/EAD-4. Argonne National Laboratory, Argonne, Ill., USA. Yu, C., et al., 2003. User's Manual for RESRAD-BUILD. Version 3, ANL/EAD/03-1. Argonne National Laboratory, Argonne, Ill., USA. Yu, C., et al., 2006. Benchmarking of RESRAD-OFFSITE: Transition from RESRAD (onsite) to RESRAD-OFFSITE and Comparison of RESRAD-OFFSITE Predictions with Peer Codes. ANL/EVS/TM/05-1. Argonne National Laboratory, Argonne, Ill., USA.
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
The potential radiological impact from a Brazilian phosphate facility cio Glo ria dos Reis*, Dejanira da Costa Lauria Ro ~o e Dosimetria (IRD/CNEN), Avenida Salvador Allende, s/n, 22780-160 Rio de Janeiro, RJ, Brazil Instituto de Radioproteça
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 August 2013 Received in revised form 6 May 2014 Accepted 3 June 2014 Available online 24 June 2014
In the semiarid region of Brazil, a facility for the production of phosphoric acid for fertilizer is in the last stages of the planning phase. The raw feedstock of Santa Quiteria has a very high level of uranium associated with the phosphate in form of apatite. The reaction by which phosphoric acid is produced generates phosphogypsum (PG) as a by-product. The ratio of phosphogypsum to phosphoric acid is approximately 5 to 1. After all of the phosphate has been extracted and processed, it is expected that some 37 million tons of phosphogypsum will be produced, containing 13 Bq/g of 226Ra and 11 Bq/g of 210 Pb. To assess the potential impact of this PG stack on the surrounding inhabitants, a generic impact assessment was performed using a modeling approach. We estimated the amount and shape of the residue stack and used computational codes for assessing the radiological impact in a prospective risk assessment. A hypothetical farmer scenario was used to calculate two potential doses, one near the site boundary and another directly over the stack piles after the project is shut down. Using a conservative approach, the potential public dose was estimated to be 2.8 mSv/y. This study identified the rainfall erosion index, dissolution rate of PG, radionuclide distribution coefficients and fish consumption rate as parameters where improved information could enhance the quality of the dose assessment. The disposal and shape of the stack is of major concern, since the PG erosion might be the main pathway for the environmental contamination; therefore, studies should be carried out to determine a suitable shape and disposal of the stack. Furthermore, containment barriers should be evaluated for their potential to reduce or avoid environmental contamination by runoff. In addition, the onsite public dose underscores the importance of a planning for remediation of the area after the plant is shut down to assure that neither the public nor the environmental health will be affected by the presence of the PG stack. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Phosphogypsum Radiological impact Modeling NORM
1. Introduction Impact assessment is a process of identifying the consequences of a past, planned or current activity. This process has been divided in two types: retrospective and prospective assessments. A retrospective assessment quantifies the impact of an ongoing activity or tries to understand the impact of a past activity, while a prospective assessment should be developed during the planning phase of a project. The overarching objective of an impact assessment is to derive a mitigation plan to minimize or eliminate negative impacts while maximizing positive impacts. It is a requirement for licensing a new facility in many countries. For the nuclear industry, a radiological impact assessment is required in addition to the conventional impact assessment and should be carried out for all phases of a project. This assessment
* Corresponding author. Tel.: þ55 21 2173 2804; fax: þ55 21 2173 2701. ria dos Reis), E-mail addresses: [email protected], [email protected] (R. Glo [email protected] (D. da Costa Lauria). http://dx.doi.org/10.1016/j.jenvrad.2014.06.001 0265-931X/© 2014 Elsevier Ltd. All rights reserved.
estimates the consequences to humans of releasing radioactive isotopes to the environment. The process is multidisciplinary and requires several areas of expertise to identify and quantify the source term, to predict the environmental transport and to estimate the human exposure to radiation and the resulting total dose, which comprises both the internal and external dose (Till and Grogan, 2008). The estimated human dose value is then compared with the adopted value of the dose limit. The level of radionuclide discharge to environment is controlled so that the dose limit is not exceeded. Any manipulation of materials that exceed the dose limit need to be controlled by a regulatory body. Apart from the nuclear industry, only industries that generate Naturally Occurring Radioactive Material (NORM) have to perform radiological impact assessments. The need to perform these assessments is a challenge for NORM industries, which must comply with standards from nuclear industries in addition to the usual environmental and worker safety requirements. A list of industry sectors have been identified as the most likely to produce NORM: oil and gas production, the niobium and ironeniobium industry, manufacture of titanium dioxide pigments, the
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
phosphate industry, the zircon and zirconia industry, production of tin, copper, aluminum, zinc, lead, and iron and steel, combustion of coal and water treatment (IAEA, 2006). The risk level for the public and the environment that can be attributed to these industries is not yet well known, not only because they differ from each other but also because the quantity of radionuclides released to the environment or contained in process residues depends on the original levels in the feedstock minerals and on the process used by the industry (UNSCEAR, 2008, IAEA, 2012). Among the NORM industries, the phosphate fertilizer industry plays an important role in conventional and radioactive pollution. Phosphate rock is used worldwide for manufacturing phosphoric acid and chemical fertilizers. It is well known that this type of rock may contain elevated levels of uranium and its decay products, while its thorium and potassium concentrations are comparable to those usually found in soils. Mortvedt (1994) reported uranium concentrations in phosphate rock worldwide ranging from 3 to 400 mg/kg, or 37e4900 Bq/kg. For 226Ra, a range from 100 to 10,000 Bq/kg has been reported (Rossler et al., 1979). The chemical process for producing phosphoric acid redistributes the radioactive isotopes among the products and residues in such way that around 90% of the U will usually be found in the phosphoric acid fraction. In contrast, 90e80% of the radium isotopes 210Pb and 210Po will be transferred from rock phosphate raw material into the phosphogypsum by-product (Rutherford et al., 1995; Hul and Burnet, 1996; Mazzili et al., 2000). Mazzili et al. (2000) reported that levels in the phosphogypsum varied between 22 and 695 Bq/kg for 226Ra, between 47 and 894 Bq/kg for 210Pb, between 53 and 677 Bq/kg for 210 Po and between 7 and 175 Bq/kg for 232Th. Therefore, radioactive isotope content of the products and residues of fertilizer production can affect not only the worker health but also the environment. The International Agency of Energy Atomic (IAEA) suggests that control and regulation of activities leading to exposure to NORM is not necessary if the effective radiation dose received by a worker or member of the public does not exceed approximately 1 mSv in a year. The IAEA (2005b) also establishes the activity concentration levels in specific materials that do not require any regulation of the related activities (1 Bq/g for uranium and thorium series radionuclides and 10 Bq/g for 40K) (IAEA, 2011). In Brazil a standard for regulating the mining and milling industries establishes a criterion for classification of the NORM industry according to reference levels of concentration varying between 10 and 500 Bq/g and a dose increment of 1 mSv/y over natural background exposure. According to this standard, the severity of regulatory control for waste management and environmental protection increases with the increase of the radionuclide concentrations in material and the resulting doses. At the moment this standard is under review (CNEN, 2005). This research supports a radiological impact assessment of the Santa Quiteria fertilizer industry (originally known as Itataia mine), which is still in its planning phase. Using a computational code that was developed for assessing the environment impact of contaminated sites, this study aims to estimate the concentrations of isotope radionuclides in various environmental media, identify the main pathway responsible for the contamination and discuss actions to minimize the potential contamination.
Based on exposure scenarios, the exposure pathways are identified. Once scenarios are defined and exposure pathways identified, a basic conceptual understanding of the system is developed and a method to estimate the dose should be chosen. For the assessment of dose, suitable environmental parameters should be identified and selected, based on site specific data whenever possible. To reduce uncertainty about the pathways and parameters, a sensitivity analysis should be performed, which may lead to acquiring more data about the site. 2.1. Source term characterization e overview of Santa Quit eria's process The Santa Quiteria raw feedstock has a very high level of uranium associated with phosphate in the form of apatite. The deposit is estimated to have 80 million tons of ore resources, with 8.9 million tons of phosphate and 62.9 thousand tons of uranium oxide s, 1984; Alc^ (IPT, 1981; Nuclebra antara e Silva and Rosa, 1988; Fukuma, 1999). According to Saad (1995) and Saad and Chamon ria project will mine 1.7 million metric (2002), the Santa Quite tons of raw material that will produce 1.5 million tons/year of phosphate rock. ria's beneficiation consists of The procedure of Santa Quite breaking up the ore by drilling and blasting, crushing the rock, and then transporting the milled ore to the beneficiation plant to separate sand and clay and to remove impurities (Fig. 1). In the plant, the material is treated by a process called desliming, where hydrocyclone equipment is used to separate the fluid mixture into two components of differing densities. From this step, an estimated 1.4 million tons of finer particles containing 10.2 Bq/g of 238U will be generated (Fukuma, 1999). Afterwards, the phosphate matrix is further upgraded by a flotation process that uses a
2. Methodology In a radiological environmental impact assessment, the first step involves gathering and evaluating existing data and information about the site, including the source term, which should be identified and characterized both qualitatively and quantitatively. Next, taking into consideration the land use, diet and habits of the inhabitants of the region, exposure scenarios should be identified.
189
Fig. 1. Flow chart of the process.
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
190
turbulent water system, chemicals and air bubbles to float the phosphate particles to the water surface, where they are skimmed off. Reis Junior et al. (1987) have estimated the content of U3O8 in this waste as 297 mg/kg, which corresponds to a 238U concentration of 3 Bq 238U/g of material. Following these steps, the beneficiated rock is transferred directly to the chemical processing plant. The production of phosria will be carried out by the wet process phoric acid in Santa Quite method that includes digestion, filtration and concentration. Sulfuric acid is used for the feedstock decomposition. By this method, the beneficiated phosphate rock is dissolved in sulfuric acid to chemically digest the calcium phosphate. This step produces a slurry that is composed of phosphoric acid solution and a suspended solid, calcium sulfate, commonly known as phosphogypsum (PG). The slurry from this digestion is filtered and the solid phase is collected, washed, and sent to phosphogypsum stacks. The phosphoric acid produced will contain most of the uranium and a separation process will be used to split uranium from the acid phase. According to Fukuma (1999), for each ton of phosphoric acid ria, approximately 4.2 tons of phosphoproduced in Santa Quite gypsum will also be produced. Accordingly, if all 8.9 million tons of P2O5 is processed, then 37 million tons of phosphogypsum will be produced. The 226Ra and 210Pb contents of this phosphogypsum are estimated to be 13 Bq/g and 11 Bq/g, respectively (Fukuma, 1999). Therefore, three main processed materials will be produced and stored at the SQ site: a) 1.4 million tons of clay that contain 10.2 Bq 238U/g; b) The flotation tailings with a 238U level of 3 Bq/g; c) 37 million tons of phosphogypsum containing 13 Bq/g of 226 Ra and 11 Bq/g of 210Pb. For this research, we selected PG as the source term because of its quantity and high radionuclide concentration. 2.2. Study area ria site is located in the northeastern region of The Santa Quite Brazil, 4 340 2700 S, 39 470 1900 W, approximately 210 km from Fortaleza city. It consists of a 40.4 km2 area that is 6250 km long and 10,000 km wide (Fig. 2).
The Semiarid climate is warm with a mean annual temperature of 27 C (varying between 21 and 33 C). Rainfall is scarce but often of high intensity; more rainfall occurs between January and April, and there is a regular drought between May and December, in which the rainfall is frequently zero. The literature reports annual mean precipitation of approximately 824 mm for the period between 1984 and 1988. Most rainfall occurs in a very short period within one month. Therefore, runoff is highly seasonal. For the period between 1984 and 1988, the reported real mean of annual gross evapotranspiration was approximately 653 mm (Lopes Filho, 1977; Silva, 2004). The site topography is characterized by surface relief and high surface elevations. The Sertaneja depression and the Residual plateaus are the two major morphological units. The ore deposit is a recessed compartment embedded between the plateaus. It has an altitude that ranges between 350 and 580 m, whereas the residual u hill is plateaus have altitudes ranging from 650 to 1000 m. The Ce the highest peak in the region with an altitude of 1085 m. This hill forms the watershed between the Curu river and Groaíras river catchments (Silva, 2004). The small rivers that drain the site are not perennial, and rivers sustain flow only during rainy season. A dam named Quixaba with a total storage capacity of 2,300,000 cubic meters is located in the site. The vegetation reflects the differences in the altitude levels. On the high altitude areas, the vegetation is remnant dry forest, and the typical vegetation of the low altitude area is savanna (Gonçalves Junior and Souza, 2012). The soils of the area are mainly from the amendment of granites and gneisses “in situ” and the colluvial materials mobilized from these rocks (CPRM, 2008). The soil is predominantly an association of Red-Yellow, Alfisols and Entisols with a low permeability (Gonçalves Junior and Souza, 2012). The common feature of the various soils are their thinness, almost always less than 1 m, that the surface horizons are uniformly sandy and that the sotopostos horizons are all sandy-clay or clay (CPRM, 2008). The groundwater flows from the south to north-northeast (Saad, 1995). The water level stays between 5 and 60 m, averaging 30 m throughout the area. The recharge of the area depends entirely on rainfall that flows on the surface and subsurface. Two main factors influence the flux: the system of open sub-vertical fractures and the topography (Silva, 2004).
ria Site. Fig. 2. Location of the PG stack and the Critical Group in the Santa Quite
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
The area surrounding the site is sparsely populated and relaria, 45 km away. The tively remote; the nearest town is Santa Quite population's economic activities are limited to scattered subsistence farms that are used for limited agricultural activities (e.g., cultivation of beans, sugar cane, corn and manioc). Much of the area is used to graze cattle and sheep. The groundwater in the area is used to supply drinking water and small scale irrigation for agriculture, whereas the surface water supports cattle watering. 2.3. Definition of exposure scenarios An exposure scenario consists of a set of human activity patterns that are related to a range of exposure pathways and include an assumed duration for each pathway. The scenarios should encompass all typical exposure situations. Therefore, each scenario represents a range of exposure situations and encompasses a set of parameter values. Through models that represent the behavior and transfer of radioactive isotopes in the environment, these scenarios relate the activity concentrations in the material to individual doses (IAEA, 2005a). The farmer scenario is the most conservative scenario, in which the land use is intensive and all exposure pathways are considered. Since this scenario attempts to consider all possible exposure pathways, this simulation produces the highest values for radiation dose. In the farmer scenario, people might be assumed to live and work on the site, grow their own food on the site, and eat fish and drink water from the local water body. ria, we developed two scenarios for the For Santa Quite simulation: i) A farmer scenario where people live close to the borders of the site (Fig. 2). According to the land use, wells supply their drinking water and the crops and animal products are locally produced. The water supply for cattle is from the dam, and the irrigation water is from wells. The worst case scenario was simulated, whereby the groundwater flows from the PG stack toward the farm and the preferential direction of the wind is also from the stack toward the farms. The exposure pathways considered are direct external gamma radiation, radon and dust inhalation, drinking water and vegetable, milk, meat, aquatic food and soil ingestion. ii) Occupation of the site by the local inhabitants after the closure of the site and subsequent abandonment. The local people are living on the top of the PG stack. This is also a farm scenario, where the exposure pathways are external gamma, radon and dust inhalation, drinking water and plant, milk, meat, aquatic food and soil ingestion.
2.4. The model for estimating the dose-RESRAD codes The chosen model for estimating dose was the pathway analysis by RESRAD codes. RESRAD is a multiple pathway analysis computer code, comprising a set of mathematic models that attempt to estimate the radiation dose and risk for an individual located either onsite or outside the contaminated zone. The contaminated zone can encompass both surface and near surface buried waste. The RESRAD codes were designed for evaluation of radiological doses to a receptor from exposure to RESidual RADioactive materials in soil (Yu et al., 1993, 2001, 2003, 2006). The RESRAD 6.5 code (onsite) is able to estimate the radiation dose of an individual who spends time directly on the primary contamination (onsite), could consume crops and animal products that are grown on the contaminated site, and drinks water from well or/and surface water located on the contaminated site. On the other hand, RESRAD offsite is able to estimate the dose of an individual located away from
191
Table 1 Parameters used for the assessments. Parameter
Value
Reference
Hydraulic conductivity for saturated soil Hydraulic conductivity for unsaturated soil Hydraulic conductivity for phosphogypsum Dry bulk density of PG Unsaturated zone thickness Gross evapotranspiration Mean annual rainfall
24.2 cm/h ¼ 2120 m/y
Ferreira et al., 1999
1.6 103 m/h (13.9 m/y) 220 cm/d ¼ 803 m/y
Local data, (Silva, 1997, 2004) Van Alphen et al. (1971)
1.15 g/cm3 30 m
Van Alphen et al. (1971) Local data, (Silva, 1997, 2004) Local data, (Silva, 1997, 2004) Local data, (Silva, 1997, 2004) Fukuma, 1999.
Radionuclide concentrations in PG
Amount of PG in the stack Shape of stack
653 mm/y 824 mm 210
Pb ¼ 210Po ¼ 11 Bq/g; Ra ¼ 13 Bq/g; 228 Ra ¼ 0.41 Bq/g 228 Th ¼ 232Th ¼ 0.026 Bq/g; 230 Th ¼ 1 Bq/g; 234 U ¼ 238U ¼ 0.4 Bq/g 37 million tons 226
3400 m 500 m wide 50 m high
Saturated zone thickness Wind velocity
200 m 2.46 m/s
PG porosity Temperature
0.7 27 C
Fukuma, 1999; INB, 2013 Estimated based on Kedainiai's stark (Kardasecicius, 2006) and Kleinschmidt, 1992. Local data, (Silva, 1997, 2004) Estimated based on Amarante et al. (2001) ~ as et al. (2007) Duen Local data, (Silva, 1997, 2004)
the primary contamination (offsite). This individual could consume vegetables and animal products from agricultural fields that are contaminated by material from the primary contamination. Likewise, his/her drinking water can be from contaminated well or surface water body located away of the contaminated site. In addition to the radiation dose and risk, the radionuclide concentration in the environmental media and soil cleanup guidelines are also determined by the codes. For the assessments, the codes consider all the major exposure pathways by which an individual may be exposed: direct external radiation; inhalation of dust and radon; and ingestion of vegetables, meat, milk, aquatic food, and soil. To estimate the contamination of offsite locations, the code models the transport from the primary contamination to agricultural areas, pastures, a dwelling area, a well, and a surface water body. The main movements of the contaminants as considered by the code are the following: i) Leaching of radionuclide caused by water that flows down through the primary contamination and causes the contamination of the groundwater. The levels of radionuclides in groundwater and their dispersion are estimated by a rate-controlled transport and a three-dimensional dispersion groundwater flow models; ii) The atmospheric dispersion of dust as modeled by equilibrium models (Plume-rise and Gaussian plume models), and iii) erosion of contaminated material by surface runoff, which is modeled as a release to the surface water body. RESRAD is able to perform sensitivity and probabilistic analysis to evaluate the influence of input parameters and generate plots. Age-specific dose factors are included in the code's database. A set of more than ninety hydrogeological, meteorological, geochemical, geometrical (size, area, depth), and material-related (soil, concrete) default parameter values are provided in the RESRAD code (Yu et al., 1993a,b). According to the authors, these
192
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
default values have been carefully selected and are realistic, although conservative, and should be used for generic modeling. For more accurate use of the code, site-specific data should be used whenever possible. The codes are user friendly, available free of charge and can be downloaded from the RESRAD Web site (http://www.evs.anl.gov/ resrad). 2.5. Parameters and input data For the methodology herein adopted, although very important the choice of parameters is not a determining factor, since the use of sensitivity analysis should point those key parameters that have most influence on the dose and should be better studied or determined. A survey of the literature was performed to set local parameters. Hence, some of the RESRAD default parameter values were changed according to the available local parameters or to parameters that best described the local characteristics (Table 1). Consequently, hydraulic conductivity, density and porosity values for PG from the literature were introduced to reflect PG characteristics. The estimated shape of the stack of 37,000,000 tons of PG was based on the PG stack in Kedainiai, which encompasses an area of 84 ha and is 60 m high, and holds 22,000,000 tons of PG (Van ~ as et al., 2007; Kardasecicius, 2006). The Alphen et al., 1971; Duen concentrations of the radionuclide daughter of short half-life were assumed to be in secular equilibrium with its parent. 3. Results and discussion 3.1. Assessment of the dose for the farm scenario located closest to the site Having taken into account all exposure pathways, the simulation using the RESRAD Offsite code for the farm located close to the site reaches a dose value for the critical group dose of about 2.5 mSv/y at 10 years and rising 2.8 mSv/y at 70 years (Fig. 3). The main contributor to this dose is 226Ra, which is responsible for 91% of the dose, followed by 210Pb. The other radionuclides did not contribute significantly to the dose (Fig. 3, Table 2). Taking into consideration the transport of contamination by environmental media, the simulation found the water-dependent pathways to be responsible for most of the dose (92%), and among those, release from the source to surface water is the predominant contamination pathway. No significant transport of contamination through groundwater was predicted by the model. The atmospheric pathway contributes 7% to the total dose, mainly due to the release of radon from the PG stack, followed by its
Fig. 3. Contribution of the radionuclides for the total dose, considering all exposure pathways.
Table 2 Total Contribution for individual radionuclides, t ¼ 70 years. Radionuclide
Total dose (mSv/y)
210
0.25 (9%) NR 2.45 (90%) 2.44 E06 1.89 E15 5.77 E03 3.97 E03 1.2 E04 1.18 E04
Pb 210 Po 226 Ra 228 Ra 228 Th 230 Th 232 Th 234 U 238 U NR ¼ not reported.
inhalation. Therefore, dust dispersion by wind was not signaled as an important contamination pathway. Among the exposure pathways, consumption of fish made the largest contribution to the total dose (86%), followed by radon inhalation, which contributed 7%. All together, the other pathways were responsible for less than 7% of the total dose (Fig. 4 and Table 2). Sensitivity analysis is a technique used to identify sensitive parameters, e.g., those having the greatest impact on dose assessment results, and to explore potential reductions in the uncertainties associated with dose parameters. The sensitive analysis, which was carried out for all parameters, noted that the rainfall erosion index was the main parameter influencing the dose. A twofold variation in the parameter value doubled or tripled the dose value (Fig. 5). Hence, according to the model, the erosion caused by rainfall might be responsible for the contamination of the surface water and consequently for the fish contamination. Based on this result, the rainfall erosion index is a parameter that should be set based on reliable site specific information data. If surface water contamination is predicted to be the major secondary source of contamination, some parameters related to the PG material dissolution in water and the radionuclide sorption by sediments might also be determined experimentally. Considering the rainfall regime in the SQ region (intense rain in a short period of time), the low permeability of soil and the scarce vegetation of the site, the erosion of material and its displacement into surface water might be a real possibility. Also the dissolution of the material in body water is a real possibility, in view of the solubility coefficients of different forms of calcium sulfate in phosphogypsum determined by Santos, who found that 2.8 g of phosphogypsum would be completely solubilized in 2000 L of water (Santos, 2002). The contamination of surface water might lead to contamination of sediment. Although sediment is not a pathway for human in the herein considered scenarios, the scenarios may change in the
Fig. 4. Contribution of the exposure pathway for the total dose.
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
Fig. 5. Influence of the erosion index caused by rainfall on the dose value.
future and sediment might be used for agricultural and building purposes for example (Rochedo et al., 2002). Also the contamination of water body may affect the local biota and an assessment of biota dose should be performed. Based on the results, the disposal and shape of the PG stack is of major concern because runoff might be the main pathway for surface water contamination. 3.2. Assessment of dose for the occupancy scenario after the closure and abandonment of the site For the scenario that considers occupancy of the site after the closure of the facility, the total dose reached the value of 160 mSv/ year at 70 years after the closure of the site, Fig. 6. In this case, the main contributor for the dose is 226Ra (222Rn), followed by 210Pb. Inhalation of radon contributes most of the total dose. Our results lead to clear conclusions. According to the current recommendations by IAEA and to Brazilian standards, the activity concentrations of the PG produced will exceed the nuclide specific exemption values of 1 Bq/g. The dose criterion for the public of 1 mSv/y is also surpassed. For that reason, the use of this material is not unrestricted, and the material would not be exempt from regulatory control. Therefore, it is unlikely that the PG produced could be used as amendment in agriculture or for construction. After the closure of this facility, the legacy waste will require strict control of the material into the foreseeable future. The total costs related to the future management should be considered in the feasibility study for the facility. Therefore, alternatives for the use and production of the PG should be considered to minimize the impact and cost of waste management. 4. Conclusion In the exposure scenarios, we considered PG as the only source term of radiation, and even so, the critical group dose was found to
193
exceed the public dose limit of 1 mSv/year (as established by the CNEN NN-3.01 regulation known as the Brazilian Basic Radiation Protection Guidelines, which is based on IAEA standards). Based on these results, the disposal and shape of the stack is of major concern because the erosion caused by rainfall may be the primary pathway for environmental contamination. Therefore, studies should be carried out to determine the parameters related to material erosion and the shape and disposal of the stack. Additionally, containment barriers should be evaluated for their potential to reduce or avoid environmental contamination by runoff. The second, or post-operational, scenario reveals a high dose for inhabitants who settle on the PG stack in the case of site abandonment. This underscores the need to plan the decommissioning of the site from the planning phase, to include this subject in the feasibility study for the project and to examine alternatives for the safe disposal or recycling of PG. In summary, the high doses under all scenarios in this study emphasize the importance of conducting an environmental risk assessment before a facility begins to operate. Because the area may eventually be reused after the closure of the facility, it is also necessary to plan for waste management to prevent potential liabilities. As IAEA has been be emphasizing the economic and social costs of remediation are incomparably higher than prevention, the results of this research point out the needs for the owner of the industry avoid possible future costs that can be unsustainable. Planning for the decommissioning of the area must occur in the initial design phase, before the construction of the facility, because measures to minimize risk and liability must be evaluated and their related costs included. It is crucial to plan for all possible future liabilities when considering the future decommissioning of the ria facility. Santa Quite The data generated and presented in this simulation are based on a conservative approach. For example, some parameters were the RESRAD defaults, and their uncertainties were not considered. The simulation noted the importance of the water pathway for environmental contamination. In light of its importance and to decrease the uncertainty in the results, site specific parameters related to this pathway should be determined whenever possible. Nevertheless, this methodology is useful for estimating risks to the public and workers in the decision-making process related to the planning and sustainability of the project. This work shows the feasibility of using a code that is designed to assess the impact of residual contamination proactively to assess the impact assessment of a facility in the planning phase of the project. Acknowledgements The authors would like to thank the CNPq (National Research Council) Brazilian funding agency for its financial support. The authors gratefully acknowledge the IAEA for promoting the improvement of skills in the field of assessment of environmental radiological impact through its EMRAS and MODARIA programmes, in which the authors are involved. Participation in these programmes has been of great importance for accomplishing this work. References
Fig. 6. Contribution for the total dose of the different exposure pathways.
^ntara e Silva, J.R., Rosa, C.V., 1988, Caracterizaça ~o geote cnica, geolo gica, química Alca rio te cnico específico. Nuclebra s, 131p. e reservas da jazida de Itataia: relato lico Brasileiro, Amarante, O.A.C., Brower, M., Zack, J., 2001. Atlas do Potencial Eo Ministerio das Minas e Energia. Brasilia, Brasil. http://www.cresesb.cepel.br/ publicacoes/download/atlas_eolico/Atlas%20do%20Potencial%20Eolico% 20Brasileiro.pdf. ~o Nacional de Energia Nuclear, 2005. CNEN-NN-4.01, Requisitos de SegurComissa ~o Radiolo gica para Instalaço ~es Minero-Industriais, Brasil. http:// ança e Proteça www.cnen.gov.br/seguranca/normas/pdf/Nrm401.pdf.
194
ria dos Reis, D. da Costa Lauria / Journal of Environmental Radioactivity 136 (2014) 188e194 R. Glo
sicas de Comiss~ ao Nacional de Energia Nuclear, 2011. CNEN-NN-3.01, Diretrizes Ba ~o Radiolo gica, Brasil. http://www.cnen.gov.br/seguranca/normas/pdf/ Proteça Nrm301.pdf. rio de Minas e Energia, Secretaria CPRM, 2008. Programa Geologia do Brasil. Ministe ~o Mineral CPRM e Serviço Geolo gico do de Geologia, Mineraç~ ao e Transformaça Brasil. Contrato CPRM- UFC N . 058/PR/05. ~ as, C., Liger, E., Can ~ ete, S., Pe rez, M., Bolívar, J.P., 2007. Exhalation of (222)Rn Duen from phosphogypsum piles located at the Southwest of Spain. J. Environ. Radioact. 95 (2e3), 63e74. ^ncia da Mineralogia da Fraça ~o Ferreira, M.M., Fernandes, B., Curi, N., 1999. Influe ~o Sudeste do Brasil. Rev. Argila nas Propriedades Físicas de Latossolos da Regia ^ncia Solo 23, 515e524. Brasil. Cie ~o de Radionuclideos Naturais nos Produtos Obtidos Fukuma, H.T., 1999. Distribuiça ~o de Acido rico a Partir da Rocha Fosfa tica de Itataia e Cear na Fabricaça Fosfo a. MSc. Thesis. Universidade de S~ ao Paulo, Brazil. ~o ambiental de Santa Gonçalves Júnior, J.T., Souza, M.J.N., 2012. Caracterizaça ria, Ceara : a nova cidade Uranífera do Brasil. Rev. Geonorte Ediça ~o Especial Quite 2 (4), 1368e1377. Hul, C.D., Burnet, W.C., 1996. Radiochemistry of Florida phosphogypsum. J. Environ. Radioact. 32, 213e238. IAEA, 2012. Environmental modelling for radiation safety (EMRAS). A summary report of the results of the EMRAS programme (2003e2007). IAEA-TECDOC1678. http://www-pub.iaea.org/MTCD/publications/PDF/TE_1678_Web.pdf. IAEA, 2005a. Guidebook on Environmental Impact Assessment for In Situ each Mining Projects. IAEA-TECDOC-1428, p. 163. IAEA, 2005b, Derivation of activity concentration values for exclusion, exemption and clearance, safety reports series No. 44. IAEA, 2006. Assessing the Need for Radiation Protection Measures in Work Involving Minerals and Raw Materials, Safety Series, vol. 49. IAEA, Vienna, p. 55. IAEA, 2011. International basic safety standards for protection against ionizing radiation and for the safety of radiation sources. Interim edition. http://wwwpub.iaea.org/MTCD/publications/PDF/p1531interim_web.pdf. Industrias Nucleares do Brasil e INB, 2013. http://www.inb.gov.br/pt-br/WebForms/ interna.aspx?secao_id¼98. gicas do Estado de Sa ~o Paulo (IPT), 1981. Relato rio n. Instituto de Pesquisas Tecnolo gico-geote cnicos para projetos de engenharia na Mina de 15-253. Estudos geolo ria/CE, vol. 1. INB, S~ Itataia: município de Santa Quite ao Paulo. Kardasecicius, L., 2006. Treakearth. http://www.trekearth.com/gallery/Europe/ Lithuania/South/Kauno/Kedainiai/photo433937.htm. Kleinschmidt, D., 1992. Phosphogypsum stack closure. In: Proceedings of the Workshop Phosphate Fertilizers and the Environment, pp. 185e192. USA. gicos para a regia ~o de Itatira. Lopes Filho, A.F., 1977. Bases de estudos hidrogeolo cnica. Nuclebra s/DRM, Fortaleza. Nota te Mazzili, B., Palmiro, V., Saueia, C., Nisti, M.B., 2000. Radiochemical characterization of Brazilian phosphogypsum. J. Environ. Radioact. 49, 113e122. Mortvedt, J.J., 1994. Plant and soil relationships of uranium and thorium decay series radio e nuclides e a review. J. Environ. Qual. 23, 643e650. 1984. Jazida de Itataia: relato rio de pesquisa mineral. INB, Fortaleza, NUCLEBRAS, p. 330.
rio Reis Junior, J.B., Aquino, J.A., Luz, I.L.O., 1987. Estudo em Planta Piloto do Mine sforo-uranífero de Itataia e CE. NUCLEBRAS/CDTN. Fo Rochedo, E.R.R., Amaral, E.C.S., Hacon, S.S., 2002. Uncertainties in the definition of critical groups to chronic exposure scenarios. In: 5th International Conference on High Levels of Natural Radiation and Radon Areas: Radiation Dose and Health Effects, Munich. Excerpta Mediaca e International Congress Proceeding Series, vol. 1225. Elsevier Science B.V, Amsterdam, pp. 25e31, 2000. Rossler, C.E., Smith, Z.A., Bolch, W.E., Prince, R.J., 1979. Uranium and radium-226 in Florida phosphate materials. J. Health Phys. 37, 269e277. Rutherford, P.M., Dudas, M.J., Arocena, J.M., 1995. Radioactivity and elemental composition of phosphogypsum produced from three phosphate rock sources. Waste Manag. Res. 13, 407e423. gico na Indústria de Fertilizantes Saad, S., 1995. Alguns Aspectos do Impacto Radiolo Fosfatados e Estudo de um Caso-Itataia (Ce). MSc. Thesis. Instituto de Engenharia Militar, Brasil. todo eletromagne tico Saad, S., Chamon, N., 2002. Levantamento geofísico com o me : (VLF) e eletrorresistividade na area da jazida de Itataia do Estado do Ceara rio de prospecça ~o geofísica. Texto, anexos e Mapas. ENGERAD, Rio de relato Janeiro, p. 86. gico ambiental do fosfogesso Santos, A.J.G., 2002, Avaliaç~ ao do impacto radiolo brasileiro e lixiviaç~ ao de 226Ra e 210Pb. 2002. 165 f. Tese (Doutorado) e ticas e Nucleares, Sa ~o Paulo, Brasil. Instituto de Pesquisas Energe ^mico- Ambientais do Uso do Fosfogesso na Silva, L.H. da C., 1997. Aspectos Econo Agricultura. MSc. thesis. Universidade Federal do Rio de Janeiro, Brasil. ~o Hidrogeolo gica da Jazida de Itataia-CE, XIII Silva J.R.A.E., 2004, Caracterizaça ^neas e Cuiab Congresso Brasileiro de Aguas Subterra a, Brasil. Till, J.E., Grogan, H., 2008. Radiological Risk Assessment and Environmental Analysis. Book. Oxford University Press Inc, New York, USA. United Nations Scientific Committee on the Effects of Atomic Radiation e UNSCEAR, 2008. Report to the General. Assembly, with scientific annexes. Vol. I: Report to the General Assembly, Scientific Annexes A and B. Van Alphen, J.G., De Los Rios Romero, F., 1971, Gypsiferous Soils: Notes on Their Characteristics and Management, International Institute for Land Reclamation and Improvement, Bulletin 12, Netherlands. Yu, C., Loureiro, C., Cheng, J.-J., Jones, L.G., Wang, Y.Y., Chia, Y.P., Faillace, E., 1993a. Data Collection Handbook to Support Modeling Impacts of Radioactive Material in Soil. Environmental Assessment and Information Sciences Division, Argonne Laboratory, Argonne, Illinois, USA. Yu, C., et al., 1993b. Manual for Implementing Residual Radioactive Material Guidelines using RESRAD. Version 5.0, ANL/EAD/LD-2. Argonne National Laboratory, Argonne, Illinois, USA. Yu, C., et al., 2001. User's Manual for RESRAD Version 6. ANL/EAD-4. Argonne National Laboratory, Argonne, Ill., USA. Yu, C., et al., 2003. User's Manual for RESRAD-BUILD. Version 3, ANL/EAD/03-1. Argonne National Laboratory, Argonne, Ill., USA. Yu, C., et al., 2006. Benchmarking of RESRAD-OFFSITE: Transition from RESRAD (onsite) to RESRAD-OFFSITE and Comparison of RESRAD-OFFSITE Predictions with Peer Codes. ANL/EVS/TM/05-1. Argonne National Laboratory, Argonne, Ill., USA.