- Email: [email protected]
Leaching of radionuclides from activated soil into groundwater
Journal of Environmental Radioactivity 143 (2015) 7e13
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Leaching of radionuclides from activated soil into groundwater F.P. La Torre*, M. Silari European Organization for Nuclear Research (CERN), Route de Meyrin, CH-1211 Geneva 23, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 August 2014 Received in revised form 31 January 2015 Accepted 31 January 2015 Available online 18 February 2015
Soil samples collected from the CERN site were irradiated by secondary radiation from the 400 GeV/c SPS proton beam at the H4IRRAD test area. Water samples were also irradiated at the same time. Detailed gamma spectrometry measurements and water scintillation analysis were performed to measure the radioactivity induced in the samples. FLUKA Monte Carlo simulations were performed to benchmark the induced radioactivity in the samples and to estimate the amount of tritium produced in the soil. Two leaching procedures were used and compared to quantify the radioactivity leached by water from the activated soil. The amount of tritium coming from both the soil moisture and the soil bulk was estimated. The present results are compared with literature data for the leaching of 3H and 22Na. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Induced radioactivity Leaching Tritium FLUKA Monte Carlo
1. Introduction Accelerators and experimental facilities are typically sited either underground or at grade with thick concrete walls and substantial earth berms to provide cost-effective shielding. Radiation interacts with the shielding materials generating induced radioactivity in the concrete or earth. In general, the induced radioactivity remains confined in the shield material. However, some activation may occur outside the accelerator enclosure, primarily in adjacent groundwater and soil. If the production of radioactive nuclides in the accelerator structure and in the concrete walls of the accelerator room is a concern for the personnel, the radionuclides created in the groundwater or in the earth represent a collective danger, because they may be transported into the environment. This was the case, for example, of the tritium leak accident discovered at Brookhaven National Laboratory (BNL) in 1997, when a considerable amount of tritium leached out of the concrete walls of a reactor's spent fuel pool. The laboratory analysis of water samples taken near the reactor revealed concentrations of tritium that greatly exceeded the EPA (US Environmental Protection Agency) drinking water standards. Shortly after the tritium levels were made public, a firestorm of public concern blew up and caused the contractor dismissal at the BNL (United States General Accounting Office, 1997).
* Corresponding author. Tel.: þ41 227677752. E-mail address: [email protected] (F.P. La Torre). http://dx.doi.org/10.1016/j.jenvrad.2015.01.020 0265-931X/© 2015 Elsevier Ltd. All rights reserved.
Several authors investigated the radioisotope production in earth and water, either experimentally or via Monte Carlo simulations (Nelson, 1965; Hoyer, 1968; Warren et al., 1969; Ranft and Goebel, 1970; Thomas, 1970; Stapleton and Thomas, 1972; Rindi, 1972; Borak et al., 1972; Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993; Baker et al., 1997; Tesch, 1997; Racky et al., 1998; Wehmann and Childress, 1999; Vincke and Stevenson, 2000; Rokni et al., 2000; Agosteo et al., 2005; Vogt et al., 2008). They describe and examine the most important isotopes of concern, but few studies have addressed the problem of the radionuclide migration from the activated soil to the groundwater. Remarkable exceptions are represented by the work of Borak et al. (1972) and Baker et al. (1997). They performed laboratory measurements of the production, leaching and movement of radionuclides produced in soil samples. Most of the later papers (Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993; Baker et al., 1997; Tesch, 1997; Racky et al., 1998; Wehmann and Childress, 1999; Vincke and Stevenson, 2000; Rokni et al., 2000) directly or indirectly refer to the results of Borak and Baker for the part concerning the leaching of radioactivity. This paper investigates the leaching of radioactivity in the type of soil (“molasse”) of the CERN region, and is part of the preventive studies conducted by the Laboratory to be prepared for a potential mishap. The scope of this study is to provide experimental data in order that, should an incident occur, transfer of radioactivity to the soil can be ready assessed. This is achieved by assessing the radionuclide production from secondary radiation in earth and water and by quantifying the activity concentration of the
8
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
radionuclides leaching into the groundwater. The study mainly focuses on 3H and 22Na, which are the longest-lived radionuclides and the most problematic from a radiation protection point of view. The chemical composition of the soil and the activation experiment are described. The results of the gamma spectrometry and liquid scintillation analyses are discussed. FLUKA Monte Carlo (Ferrari et al., 2005; Battistoni et al., 2007) simulations were performed to estimate the amount of tritium in the soil. In order to validate the FLUKA capability in predicting the production of radioactivity in this specific context, the simulation results were benchmarked against the measured radionuclides. Two alternative approaches were investigated: 1) the mixing system, in which water is mixed with the soil and 2) the flowing system, in which the water flows through the soil. The results of both leaching procedures are analysed and compared. The amount of tritium coming from the soil moisture and the soil bulk was also estimated, a novel result with respect to previous studies.
3. Irradiation set-up
2. Soil and water samples The soil samples used for the irradiation were collected on the vessin site of CERN, at a depth of 24 m. CERN is located in the Pre Geneva basin, which is filled by sedimentary deposits called molasse. The chemical analysis of the molasse rock was carried out by the EMPA laboratory in Dübendorf (Switzerland) via X-ray fluorescence spectrometry (WD-XRF). Soil samples were dried before the XRF measurements. The XRF analysis is not reliable in quantifying the 1st and 2nd period elements (from H to Ne) in inorganic solid matrices (EMPA, 2011). For this reason, the oxygen and carbon contents were extrapolated using typical values of molasse soil (Vincke and Stevenson, 2000). The results of this chemical assay are shown in Table 1. The knowledge of the natural water content of the soil (moisture) is essential to discriminate the fraction of radioactivity leaching into groundwater from the soil, from the fraction coming from the moisture. The moisture was measured in CERN Environmental laboratory by drying a known amount of earth. The soil moisture content may be expressed by weight as the ratio of the mass of water present to the dry weight of the soil sample. To determine this ratio, the samples and container were weighed in the laboratory both before and after drying, the difference being the mass of water originally in the sample. The water content measured in the soil specimen was 5% by weight. Three cylindrical plastic containers 5 cm long and 4.5 cm in diameter were filled with 100 g of soil. A sample of drinking water was collected in a 50 ml plastic container 10 cm long and 2.5 cm in diameter (Fig. 1). The natural radioactivity in the samples was measured before irradiating them in the experiment. A sample of molasse was measured with a Ge detector for 5 104 s (about 14 h). The main contribution of the natural radioactivity comes from 40K (0.32 Bq g1) with small contributions from the thorium (0.018 Bq g1) and uranium chains (0.019 Bq g1). These results are in good agreement with the activity concentrations given in Vincke and Stevenson (2000). Additional information on the samples and on the chemical analysis can be found in La Torre et al. (2012a).
Table 1 Chemical composition of dried soil sample (density: ~1.4 g cm3). Element
O
Si
Ca
Al
C
Fe
Mg
K
Na
(g/100 g)
38.8a
24
16
6.8
5a
4
2
1.9
0.7
0.42
Mn
Ba
P
Sr
Zn
Cr
Zr
Eu
Ni
S
0.11
0.06
0.06
0.05
0.03
0.02
0.02
0.01
0.01
0.01
(g/100 g) a
Fig. 1. Soil and water samples used in the activation experiment.
Extrapolated value, not quantifiable by XRF analysis.
Ti
The activation experiment was carried out at the H4IRRAD facility in the North Experimental Area of the CERN Super Proton Synchrotron (SPS). This facility hosts a copper target struck by the SPS primary proton beam with momentum of 400 GeV/c and average intensity of 3 109 protons per pulse (over a supercycle of about 45 s and an extraction length of ~5 s). The soil and water containers were installed under the copper target. A schematic view of the experimental set up is shown in Fig. 2. The activation experiment was carried out at the H4IRRAD facility (Biskup et al., 2011) in CERN North Experimental Area. The soil and water containers were installed under the copper target struck by the SPS primary proton beam with momentum of 400 GeV/c and average intensity of 3 109 protons per pulse (over a supercycle of about 45 s and an extraction length of ~5 s). An argon ionisation chamber (XION) placed in the H4 beam line just upstream of the copper target monitored the intensity of the primary beam. One XIONcount corresponds to (6900 ± 690) particles impinging on the target (La Torre et al., 2012b). 3.1. Analysis of the irradiated samples Immediately after irradiation with ~ 7.5 1013 accumulated protons, the dose rate of the samples was of the order of a few mSv/ h. Most of this radioactivity was due to very short half-life radioisotopes. Since the radioisotopes of interest to this study have medium and long half-life, the samples were let decaying for 10 days before counting. The earth and water samples were measured with a high sensitivity, low-background, high-purity germanium (HPGe) detector by Canberra. The data acquisition and analysis was carried out using Canberra's Genie-2000 spectrometry software and the PROcount-2000 counting procedure software (Genie-2000 Software, 2000). This is a comprehensive software package for data acquisition, display and analysis, which includes a set of advanced spectrum analysis algorithms providing a complete analysis of gamma ray spectra. Three gamma spectrometry analyses were performed for each sample, at cooling times of one week, one month and two months. The soil samples could not be directly counted for 3H due to the low beta-particle endpoint energy (18 keV), which is absorbed in the sample. For this reason, the tritium activity in the soil was estimated via Monte Carlo calculations (Section 3.2). The tritium activity in the water was determined using a liquid scintillation counter (Packard TRI-CARB 3180TR/SL), measuring a mixture of 8 ml of activated water and 12 ml of so-called liquid scintillation cocktail (Packard Ultima Gold LLF). In case of high precision
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
9
measurements, distillation is usually recommended requiring well controlled conditions where other radionuclides present in the sample (e.g. 22Na) may significantly increase the result for tritium. This was not needed in the present case, as the potential interference of other radionuclides in the tritium pulse-height window was negligible. Tables 2 and 3 show the results of the gamma spectrometry and tritium measurements for the soil and water samples after 10 days of cooling time. Only radionuclides with the smallest experimental uncertainty and comparable cooling time and half-life are listed. Quoted errors include statistical and non-statistical uncertainties of the gamma spectrometry analysis. 3.2. FLUKA simulations The specific activities of the radionuclides in the samples were predicted by Monte Carlo simulations with the FLUKA code (Ferrari et al., 2005; Battistoni et al., 2007). The simulations were performed using a detailed geometrical model of the experimental facility implemented by the H4IRRAD team (Biskup et al., 2011). Fig. 2 shows the target station as implemented in the FLUKA simulations. The irradiated samples were reproduced in details with the actual size, the Plexiglas container and their position near the target. The elemental compositions of the samples were those obtained from the chemical analysis, to which 5% of water (moisture) was added. The full hadronic cascade was simulated in the irradiation area. The nuclear spallation process is the main reaction mechanism at 400 GeV energy. Tritium is produced by spallation from oxygen atoms in water and soil. The production of 22Na is due to spallation reactions from 23Na and heavier elements in soil (e.g. 24Mg, 27Al, 28 Si). For an accurate description of all nuclear processes relevant for isotope production in the soil and water samples, the coalescence mechanism and the evaporation of heavy fragments were explicitly turned on via two separated PHYSICS cards. The
Fig. 3. Fluence spectra of protons, neutrons and charged pions at the location of the soil and water samples.
coalescence mechanism describes the emission of energetic light fragments (up to alpha particles), while the production of fragments up to mass 24 is described by the evaporation of heavy fragments. The algorithm adopts a sampling scheme for the emitted particle spectra including sub-barrier effects and takes the full energy dependence of the nuclear level densities into account. These physics models allow a much more accurate description of the production of residual nuclei (Ferrari et al., 2005). To obtain an accurate simulation of the physical process involved in the activation experiment, the card DEFAULTS was used. The particle transport threshold was set at 100 keV, except for neutrons transported down to thermal energies. Low energy neutrons (below 20 MeV) were transported using the 260 group library. A total of 3 105 primary particle histories were simulated in 100 Monte Carlo runs. The total irradiation time was 1.73 106 s with an average beam intensity of 4.33 107 protons per second, i.e. about 7.5 1013 accumulated number of protons. Region-importance biasing was used in order to enhance the statistical accuracy of the results. The production of residual nuclei and their radioactive decay were scored in the same run. The RESNUCLEi card scores the residual nuclei produced in inelastic interactions, while the radioactive decay was calculated using the RADDECAY, DCYSCORE, DCYTIMES and IRRPROFIle cards, taking into account the decay chains, build-up of isotopes and the irradiation profile. The USRTRACK card scores the particle spectral fluences in the sample region. Fig. 3 shows the differential distributions of the energy fluence of neutrons, protons and charged pions. Experimental values, FLUKA predictions and their ratios are shown in Table 2 for the soil and in Table 3 for the water sample.
Table 2 Results of experiment and calculations for the specific activity in the soil sample. Nuclide
3
H Be Na 46 Sc 48 V 51 Cr 52 Mn 54 Mn 56 Co 58 Co 7
22
Fig. 2. Irradiation set-up in the FLUKA geometry drawn with SimpleGeo (Theis et al., 2006). The aluminium target holder is shown in green, the copper target in blue and the irradiated soil sample in brown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
t1/2
12.32 y 53.1 d 2.6 y 83.8 d 16 d 27.7 d 5.59 d 312 d 77.3 d 70.9 d
Specific activity (Bq g1)
Ratio
Experimental
FLUKA
Exp/FLUKA
e 425 ± 32 12.5 ± 0.8 3.02 ± 0.19 9.33 ± 0.50 25.0 ± 2.3 6.01 ± 0.43 6.02 ± 0.42 0.40 ± 0.04 0.22 ± 0.05
14.1 257 10.2 2.99 12.3 27.2 6.95 5.74 0.86 0.35
± ± ± ± ± ± ± ± ± ±
0.3 4 0.2 0.35 1.1 1.4 0.49 0.23 0.09 0.05
e 1.65 1.22 1.01 0.76 0.92 0.86 1.05 0.46 0.63
10
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
Table 3 Results of experiment and calculations for the specific activity in the water sample. Nuclide
3 7
H Be
t1/2
12.32 y 53.1 d
Specific activity (Bq g1)
Ratio
Experimental
FLUKA
Exp/FLUKA
28.9 ± 2.6 523 ± 48
27.6 ± 0.5 480 ± 12
1.05 1.09
4. Leaching procedures Radionuclides affecting the groundwater come from leaching. Two leaching possibilities were investigated (Fig. 4): 1) water stagnation with irradiated soil (mixing system); 2) water percolation through the irradiated soil (flowing system). In this work, we used the standard leaching procedure for the measurements of the leaching factor, consisting of 10 parts by weight of water and one part of irradiated soil (Borak et al., 1972; Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993). To measure the amount of radioactivity passed from the irradiated soil to the water in both systems, leached water samples were systematically measured by gamma spectrometry and by scintillation analysis. 4.1. Soilewater mixing system In the mixing system, 100 g of the irradiated soil were placed in a graduated flask together with 1 l of distilled (tritium free) water (Fig. 4 e left). After vigorous shaking to disperse the soil in the water, the mixture was stirred for 8 h. To measure the radioactivity leached out, a 100 ml sample of the hazy water was filtered through a Millipore filter (0.45 mm). The gamma activity in the water was measured with a Germanium detector while b emitters were measured with a liquid scintillation counter after distillation.1 Activated soil and water were in contact for 4 months and the analyses on the leached water were repeated after 1, 2 and 4 months. 4.2. Soilewater flowing system In the flowing system (Fig. 4 e right) a sample of 100 g of irradiated soil was placed in a funnel with a Millipore filter (0.45 mm)
Table 4 Radioactivity measured in the leached water after correction for the decay and the concentration for the soilewater mixing system. Mixing time
Specific activity (Bq l1)
8 1 2 4
518 541 542 542
3 H (t1/2 ¼ 12.32 y)
h month months months
± ± ± ±
31 32 32 33
7
Be (t1/2 ¼ 53.1 d)
22
5.84 ± 3.03
96.7 131 136 144
Na (t1/2 ¼ 2.6 y) ± ± ± ±
5.8 8 8 9
48 V (t1/2 ¼ 16 d)
8.21 ± 0.49 10.2 ± 1.3 16.1 ± 2.1
MDA: minimum detectable activity.
connected to a graduated container. A glass separatory funnel was placed over the funnel and filled with 1 l of distilled (tritium free) water. The glass stopcock allowed controlling the rate of addition of the water to 125 ml h1. The gamma activity in the water was measured with a Germanium detector while b emitters were measured with a liquid scintillation counter after distillation. The whole procedure was repeated after 1, 2 and 4 months. 5. Data analysis The experimental results were analysed and corrected for the decay time and for the radioactivity concentration. The error bars shown in Tables 4 and 5 were estimated using the propagation of uncertainty law based on the uncertainties of the results of the gamma spectrometry and liquid scintillation measurements that include statistical and non-statistical errors. The specific activities measured in the water as a function of the mixing time for the soilewater mixing system are given in Fig. 5 and in Table 4. The 3H concentration in the water quickly increases to 95.6% (518 Bq l1) of the total activity after 8 h of mixing time. It takes one month to rise to 99.8% (541 Bq l1) and 2 months to reach 542 Bq l1, the maximum value measured in the water. A similar behaviour was found for 22Na. The highest concentration was measured after 123 days with 144 Bq l1. Eight hours are enough to reach 67% (96.7 Bq l1) of the total 22Na activity measured in the leached water. Half of the total 48V activity was measured after 8 h and the maximum activity is measured after 61 days with 16.1 Bq l1. The results of the water analysis for the soilewater flowing system are given in Fig. 6 and in Table 5. The specific activity of 3H decreases from 384 Bq l1 to 3.23 Bq l1 after the first wash. The same behaviour is observed for 22Na that passes from 126 Bq l1 to 12.7 Bq l1 in one wash. The concentration of 48V and 51Cr varies little even after three washes: 48V slowly decreases from 4.51 Bq l1 to 1.95 Bq l1 whereas 51Cr decreases from 13.0 Bq l1 to 8.33 Bq l1. 5.1. Fraction of radioactivity leached into the water The activity concentration of the leached water for both systems was compared with the radioactivity measured in the activated soil Table 5 Radioactivity measured in the leached water including the decay correction for the soilewater flowing system. Wash Specific activity (Bq l1) 3 7 22 48 51 H Be Na V Cr (t1/2 ¼ 12.32 y) (t1/2 ¼ 53.1 d) (t1/2 ¼ 2.6 y) (t1/2 ¼ 16 d) (t1/2 ¼ 27.7 d)
Fig. 4. Soilewater mixing system (left) and flowing system (right).
1 100 ml of leached water were evaporated to dryness. During this process approximately 20 ml of condensate was collected for 3H measurement.
1 2 3 4
384 3.23 1.67 2.05
± ± ± ±
23 0.93 0.90 1.09
8.48 ± 2.40
MDA: minimum detectable activity.
126 12.7 3.15 4.71
± ± ± ±
6 0.9 0.22 0.54
4.51 ± 0.28 2.55 ± 1.4 1.95 ± 1.23
13.0 ± 3.2 9.16 ± 4.57 8.33 ± 3.67
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
11
Table 6 Fraction of radioactivity leached by water in the mixing system. Nuclide t1/2
tot Leachable fraction for mixing (%) ½Atot i ðleached waterÞ=Ai ðsoilÞ
8h 3
a
H Be Na 48 V 7
22
12.32 y 36.7 ± 7.7 53.1 d 0.014 ± 0.007 2.6 y 7.74 ± 0.68 16 d 0.88 ± 0.08
1 month
2 months
4 months
38.4 ± 8.0 e 10.5 ± 0.9 1.09 ± 60.16
38.5 ± 8.1 e 10.9 ± 1.0 1.73 ± 0.44
38.5 ± 8.1 e 11.5 ± 1.1 e
a 3 H activity measured in the leached water compared with the activity estimated with FLUKA in soil.
the tritium activity measured in the activated water sample (Table 3). Since the measured soil moisture is 5% by weight (see Section 2), i.e. 5 g for 100 g of soil, it can be assumed that:
Atot trit ðmoistureÞ ¼ Atrit ðwaterÞ 5 g ¼ ð145±9Þ Bq Fig. 5. Specific activity in the leached water as a function of the mixing time for the soilewater mixing system, after correction for the decay and the concentration.
in order to estimate the fraction leached out. Tables 6 and 7 show the ratio between the total activity measured in the leached water and the total activity measured in the activated soil for each tot radionuclide ½Atot i ðleached waterÞ=Ai ðsoilÞ for the mixing and flowing systems, respectively. Since the liquid scintillation analysis cannot be performed on the soil, the tritium activity measured in the leached water was compared with the tritium estimated in the soil by the FLUKA simulations. Fig. 7 shows the leachable fraction of 3 H and 22Na, the two radionuclides of major interest in this study. Most of the radioactivity leached into the water just after 8 h of mixing time (mixing system) or after the first wash (flowing system). In the first case, the longer the soil is mixed with the water, the more radioactivity is leached out. After two months of stirring, 39% of the 3H and 11% of the 22Na is leached by the water. For the flowing system, the leachable fraction after one wash is 27% for 3H and 10% for 22Na. As from the second wash, the leachable fraction falls down to a fraction of per cent for both radioisotopes. It is clear that the mixing system is a much more efficient way to extract radioactivity from the activated soil. It is also interesting to compare the tritium extracted by water with its activity in the bulk (dried soil) and in the moisture of the activated soil. The tritium activity in the soil moisture Atot ðmoistureÞ can be estimated from trit
(1)
where Atrit(water) is the specific activity of 3H in the activated water, i.e. (28.9 ± 2.9) Bq g1. The total activity of 3H in the soil, Atot ðsoilÞ, is given by: trit
Atot trit ðsoilÞ ¼ Atrit ðsoilÞ 100 g ¼ ð1410±30Þ Bq
(2)
where Atrit(soil) is the specific activity of 3H estimated by FLUKA in the soil (Table 2), i.e. (14.1 ± 0.3) Bq g1. The tritium activity in the bulk of the activated soil Atot trit ðbulkÞ can thus be estimated as follow: tot tot Atot trit ðbulkÞ ¼ Atrit ðsoilÞ Atrit ðmoistureÞ ¼ ð1265±105Þ Bq
(3)
where Atot ðmoistureÞ and Atot ðsoilÞ are given by expressions (1) trit trit and (2). Tritium produced in soil bulk is much more difficult to extract than tritiated water moisture, which behaves chemically like water (Canadian Nuclear Safety Commission (CNSC), 2013). For this reason, it can be assumed that both techniques extracted before the tritium produced in the soil moisture so that the remaining part of tritium can be attributed to the soil bulk. Table 8 summarizes the results and shows the leachable fraction of the moisture and the leachable fraction of the bulk for the mixing and flowing systems. According to the previous assumption, it could be inferred that both systems leached out 100% of tritium produced in the moisture; the mixing system removed 32% of tritium produced in the bulk, whereas the flowing system removed only 20%. 6. Comparison with literature data Borak et al. (1972) report a series of experiments carried out to determine the most practical approach for measuring the leachability of elements from the soil (glacial till). Finally, a batch processing technique by directly mixing 10 parts by weight of water and one part of soil was used. Baker et al. (1997) tested several soil Table 7 Fraction of radioactivity leached by water in the flowing system. Nuclide
t1/2
Leachable fraction for flowing (%) tot ½Atot i ðleached waterÞ=Ai ðsoilÞ
3
12.32 y 53.1 d 2.6 y 16 d 27.7 d
27.2 0.02 10.1 0.48 0.52
1 wash Ha 7 Be 22 Na 48 V 51 Cr Fig. 6. Specific activity in the leached water as a function of wash including decay time correction for the wateresoil flowing system.
± ± ± ± ±
5.6 0.01 0.8 0.04 0.13
2 washes 0.23 e 1.02 0.27 0.37
± 0.08 ± 0.10 ± 0.14 ± 0.18
3 washes 0.12 e 0.25 0.21 0.33
± 0.07 ± 0.02 ± 0.13 ± 0.15
4 washes 0.15 ± 0.08 e 0.38 ± 0.04 e e
a 3 H activity measured in the leached water compared with the activity estimated with FLUKA in soil.
12
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13 Table 9 Comparison of the present results for 3H and
3
H
22
Na
22
Na with literature data.
This work (mixing)
This work (flowing)
Borak et al. (1972)
Baker et al. (1997)
39% (100% moisture, 31% bulk) 12%
28% (100% moisture, 20% bulk) 11%
100%
66e100%
10e20%
7e32%
components. This study was carried out for a specific type of soil, the molasse, and the results of leaching of radioactivity from the soil bulk should not be generalized to other type of soils without a dedicated investigation. 7. Conclusions
Fig. 7. Cumulative fraction of 3H and vated soil.
22
Na activities leached by water from the acti-
samples (Weathered Austin chalk, Eagle Ford shale, Ellis County soil) using the same technique of Borak et al., but mixing an amount of distilled water equal in weight to the weight of the sample. Both authors determined the amount of 3H produced in the soil by distillation of the activated soil in an oven and collecting the water driven off. This water was then assayed for 3H by means of liquid scintillation counting. Borak et al. found that the leached water accounted for all of the 3 H measured by the bake out process, on the other hand the fraction of 22Na removed from the activated soil was about 10e20%. Baker et al. measured a 3H percentage leachable in the range of 66e100% and an amount of 22Na in the leached water ranging between 7% and 32%. Table 9 compares the present results with those of Borak et al. and Baker et al. for 3H and 22Na. If the 22Na measured in the leached water is compatible with the literature data, a large discrepancy is observed for 3H. Borak et al. and Baker et al. made the assumption that all the 3H produced in the sample that could leach was transferred to the water that comes out in the distillation process. This assumption leads to an overestimation of the total 3H leachable fraction, since the evidence suggests that not all the tritium produced from soil molecules can be extracted with the distillation process. This work seems to show that 100% of the 3H produced in the moisture is measured in the leached water, but only 31% (mixing) and 20% (flowing) of 3H produced in soil bulk can leach out. It is important to mention that other parameters can influence the leaching process of radioactivity from the soil. For example, the soil composition and particularly its organic fraction can have a significant impact on the retention of dissolved radioactive
Table 8 Tritium percentage leachable from soil bulk and moisture for both systems. Activity (Bq) 3
Soil bulk (95 g) Soil moisture (5 g) Leached water Fraction of 3H moisture leached out Fraction of 3H bulk leached out
H (mixing)
1265 ± 105 145 ± 9 542 ± 33 100% 31.4%
3
H (flowing)
384 ± 23 100% 18.9%
CERN soil and water samples were irradiated in the H4IRRAD test facility. Gamma spectrometry and beta scintillation analysis were performed to determine the radioactivity induced in the samples. The main radionuclides found in the soil after 10 days of cooling time are 7Be, 22Na, 46Sc, 48V, 51Cr, 52Mn, 54Mn. The water radioactivity is due to 3H and 7Be radionuclides. FLUKA simulations were benchmarked against the experimental results and allowed estimating the concentration of 3H in the soil. Two leaching procedures were used and compared to quantify the amount of radioactivity leached out of the soil into the water. The mixing system was able to remove up to 39% of 3H and 12% of 22Na from the irradiated soil. The flowing system extracted 28% of 3H and 11% of 22 Na. The results seem to indicate that 100% of the tritium produced in the soil moisture may leach out, but only 31% (mixing) and 20% (flowing) of the tritium produced in the soil bulk can be extracted. Thus most of the transferable 3H produced in soil depends on the amount of water in the soil at the time of irradiation. On the other hand, only a small percentage (31% in mixing and 20% in flowing) of tritium, which is soil-type dependent, can leach out. This work provides previously unavailable information on the tritium leachable fraction from activated soil. Acknowledgements We wish to thank P. Vojtyla for providing access to the Environmental Laboratory and for many helpful discussions. We would like to thank A. Dziewa and F. Malacrida for performing the gamma spectrometry and the liquid scintillation measurements. We are grateful to M. Calviani, M. Brugger and the H4IRRAD team for allowing the irradiation of the samples and for providing the FLUKA geometry of the facility. References Agosteo, S., Magistris, M., Silari, M., 2005. Radiological considerations on multi-MW targets, part I: induced radioactivity. Nucl. Instrum. Methods Phys. Res. A 545, 813e822. Baker, S.I., 1975. Soil Activation Measurements at Fermi Laboratory (Internal report). Fermi National Accelerator Laboratory. Baker, S.I., 1985. Fermilab soil activation experience. In: Paper Presented at the Fifth DOE Environmental Protection Information Meeting, Albuquerque, New Mexico, November 1984, CONF-841187, pp. 673e683. Baker, S., Bull, J., Gross, D., 1997. Leaching of accelerator produced radionuclides. Health Phys. 73, 912e918. Battistoni, G., Muraro, S., Sala, P.R., Cerutti, F., Ferrari, A., Roesler, S., Fasso, A., Ranft, J., 2007. The FLUKA code: description and benchmarking. In: Albrow, M., Raja, R. (Eds.), Proceedings of the Hadronic Shower Simulation Workshop 2006, Fermilab, 6e8 September 2006, AIP Conference Proceeding, vol. 896, pp. 31e49. Biskup, B., Brugger, M., Calviani, M., Efthymiopoulos, I., Kwee, R., Mekki, J., La Torre, F.P., Lebbos, E., Mala, P., Manessi, G., Nordt, A., Pozzi, F., Roeed, K., Severino, C., Silari, M., Thornton, A., 2011. Commissioning and Operation of the H4IRRAD Mixed-field Test Area (CERN technical note, CERN-ATS-Note-2011-121 PERF).
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13 Borak, T.B., Awschalom, M., Fairman, W., Iwami, F., Sedlet, J., 1972. The underground migration of radionuclides produced in soil near high-energy protons accelerators. Health Phys. 23, 679e687. Canadian Nuclear Safety Commission (CNSC), 2013. Environmental Fate of Tritium in Soil and Vegetation, ISBN 978-1-100-22687-3. EMPA e Swiss Federal Laboratories for Material Science and Technology, 2011. Laboratory for Solid State Chemistry and Catalysis, Überlandstrasse 129, CH8600 Dübendorf, Test Report n. 459595. Ferrari, A., Sala, P.R., Fasso, A., Ranft, J., 2005. FLUKA: a Multi-particle Transport Code (CERN-2005-10, INFN/TC_05/11, SLAC-R-773). Genie-2000 Spectroscopy Software. ISO 9001. Canberra Industries. http://www. canberra.com. Hoyer, F., 1968. Induced Radioactivity in the Earth Shielding on Top of High-energy Particle Accelerators (CERN 68-42). La Torre, F.P., Magistris, M., Silari, M., 2012a. Irradiation of Soil Samples at H4IRRAD to Study Radioisotope Leaching into Groundwater: Preliminary Study (CERN technical report, CERN-DGS-2012-047-RP-TN). La Torre, F.P., Manessi, G.P., Pozzi, F., Severino, C.T., Silari, M., 2012b. Cu and Al Activation Experiments for Beam Monitoring in H4IRRAD (CERN technical report, CERN-DGS-2012-022-RP-TN). Malensek, A.J., Wehmann, A.A., Elwyn, A.J., Moss, K.J., Kesich, P.M., 1993. Groundwater Migration of Radionuclides at Fermilab (Fermilab report, FERMILAB-TM1851). Nelson, W.R., 1965. Radioactive Groundwater Produced in the Vicinity of Beam Dumps (Stanford linear accelerator, center international note SLAC-TN-65-16). Racky, B., Dinter, H., Leuschner, A., Tesch, K., 1998. Radiation environment of the linear collider TESLA. In: Proceedings of the Fourth Workshop on Simulating Accelerator Radiation Environments (SARE-4), Knoxville, Tennessee, U.S.A., p. 14. Ranft, J., Goebel, K., 1970. Estimation of Induced Radioactivity Around High-energy Accelerators from Hadronic Cascade Star Densities Obtained from Monte Carlo Calculations (CERN HP-70-92).
13
Rindi, A., 1972. Induced Radioactivity in the Cooling Waters of the CERN 300 GeV SPS (CERN international report Lab II-RA/Note/72-71). Rokni, S., Liu, J.C., Roesler, S., May 2000. Initial Estimates of the Activation Concentration of the Soil and Groundwater Around the NLC Beam Delivery System Tunnel (SLAC RP note RP-00-04). Stapleton, G.B., Thomas, R.H., 1972. Estimation of the induced radioactivity of the groundwater system in the neighbourhood of a proposed 300 GeV high-energy accelerator situated on a chalk site. Health Phys. 23, 689. Sullivan, A.H., 1987. Groundwater Activation Around the AA and ACOL Target Areas (CERN internal report CERN/TIS-RP/IR/87-34). Tesch, K., 1997. Production of Radioactive Nuclides in Soil and Groundwater Near Dump of a Linear Collider (DESY internal report-DESY, D3-D86). Theis, C., Buchegger, K.H., Brugger, M., Forkel-Wirth, D., Roesler, S., Vincke, H., 2006. Interactive three dimensional visualization and creation of geometries for Monte Carlo calculations. Nucl. Instrum. Methods Phys. Res. A 562, 827e829. Thomas, R.H., 1970. Possible Contamination of Groundwater System by High-energy Acceleration (Internal report UCRL-20131). Lawrence Berkley Laboratory. United States General Accounting Office, 1997. Information on the Tritium Leak and Contractor Dismissal at the Brookhaven National Laboratory (Department of Energy, GAO/RCED-98-26). Vincke, H., Stevenson, G.R., 2000. Production of radioactive isotopes in molasse. In: 5th Meeting of the Task Force on Shielding Aspects of Accelerators, Targets and Irradiation Facilities, Paris, France, 18e21 Jul 2000, pp. 47e57. Vogt, K., Haida, M., Fehrenbacher, G., 2008. Soil activation studies for the FAIR project. In: 1st ARIA 2008 Workshop on Accelerator Radiation Induced Activation, October 13e17, 2008. Paul Scherrer Institut (PSI), Switzerland. Warren, G.J., Busick, D.D., McCall, R.C., 1969. Radioactivity produced and released from water at high energies. In: Proc. Second Int. Conf. in Accelerator Dosimetry, Stanford, November 1969, CONF-691101. Wehmann, A., Childress, S., 1999. Tritium Production in the Dolomitic Rock Adjacent to NuMI Beam Tunnels (Fermilab report, FERMILAB-TM-2083).
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Leaching of radionuclides from activated soil into groundwater F.P. La Torre*, M. Silari European Organization for Nuclear Research (CERN), Route de Meyrin, CH-1211 Geneva 23, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 August 2014 Received in revised form 31 January 2015 Accepted 31 January 2015 Available online 18 February 2015
Soil samples collected from the CERN site were irradiated by secondary radiation from the 400 GeV/c SPS proton beam at the H4IRRAD test area. Water samples were also irradiated at the same time. Detailed gamma spectrometry measurements and water scintillation analysis were performed to measure the radioactivity induced in the samples. FLUKA Monte Carlo simulations were performed to benchmark the induced radioactivity in the samples and to estimate the amount of tritium produced in the soil. Two leaching procedures were used and compared to quantify the radioactivity leached by water from the activated soil. The amount of tritium coming from both the soil moisture and the soil bulk was estimated. The present results are compared with literature data for the leaching of 3H and 22Na. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Induced radioactivity Leaching Tritium FLUKA Monte Carlo
1. Introduction Accelerators and experimental facilities are typically sited either underground or at grade with thick concrete walls and substantial earth berms to provide cost-effective shielding. Radiation interacts with the shielding materials generating induced radioactivity in the concrete or earth. In general, the induced radioactivity remains confined in the shield material. However, some activation may occur outside the accelerator enclosure, primarily in adjacent groundwater and soil. If the production of radioactive nuclides in the accelerator structure and in the concrete walls of the accelerator room is a concern for the personnel, the radionuclides created in the groundwater or in the earth represent a collective danger, because they may be transported into the environment. This was the case, for example, of the tritium leak accident discovered at Brookhaven National Laboratory (BNL) in 1997, when a considerable amount of tritium leached out of the concrete walls of a reactor's spent fuel pool. The laboratory analysis of water samples taken near the reactor revealed concentrations of tritium that greatly exceeded the EPA (US Environmental Protection Agency) drinking water standards. Shortly after the tritium levels were made public, a firestorm of public concern blew up and caused the contractor dismissal at the BNL (United States General Accounting Office, 1997).
* Corresponding author. Tel.: þ41 227677752. E-mail address: [email protected] (F.P. La Torre). http://dx.doi.org/10.1016/j.jenvrad.2015.01.020 0265-931X/© 2015 Elsevier Ltd. All rights reserved.
Several authors investigated the radioisotope production in earth and water, either experimentally or via Monte Carlo simulations (Nelson, 1965; Hoyer, 1968; Warren et al., 1969; Ranft and Goebel, 1970; Thomas, 1970; Stapleton and Thomas, 1972; Rindi, 1972; Borak et al., 1972; Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993; Baker et al., 1997; Tesch, 1997; Racky et al., 1998; Wehmann and Childress, 1999; Vincke and Stevenson, 2000; Rokni et al., 2000; Agosteo et al., 2005; Vogt et al., 2008). They describe and examine the most important isotopes of concern, but few studies have addressed the problem of the radionuclide migration from the activated soil to the groundwater. Remarkable exceptions are represented by the work of Borak et al. (1972) and Baker et al. (1997). They performed laboratory measurements of the production, leaching and movement of radionuclides produced in soil samples. Most of the later papers (Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993; Baker et al., 1997; Tesch, 1997; Racky et al., 1998; Wehmann and Childress, 1999; Vincke and Stevenson, 2000; Rokni et al., 2000) directly or indirectly refer to the results of Borak and Baker for the part concerning the leaching of radioactivity. This paper investigates the leaching of radioactivity in the type of soil (“molasse”) of the CERN region, and is part of the preventive studies conducted by the Laboratory to be prepared for a potential mishap. The scope of this study is to provide experimental data in order that, should an incident occur, transfer of radioactivity to the soil can be ready assessed. This is achieved by assessing the radionuclide production from secondary radiation in earth and water and by quantifying the activity concentration of the
8
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
radionuclides leaching into the groundwater. The study mainly focuses on 3H and 22Na, which are the longest-lived radionuclides and the most problematic from a radiation protection point of view. The chemical composition of the soil and the activation experiment are described. The results of the gamma spectrometry and liquid scintillation analyses are discussed. FLUKA Monte Carlo (Ferrari et al., 2005; Battistoni et al., 2007) simulations were performed to estimate the amount of tritium in the soil. In order to validate the FLUKA capability in predicting the production of radioactivity in this specific context, the simulation results were benchmarked against the measured radionuclides. Two alternative approaches were investigated: 1) the mixing system, in which water is mixed with the soil and 2) the flowing system, in which the water flows through the soil. The results of both leaching procedures are analysed and compared. The amount of tritium coming from the soil moisture and the soil bulk was also estimated, a novel result with respect to previous studies.
3. Irradiation set-up
2. Soil and water samples The soil samples used for the irradiation were collected on the vessin site of CERN, at a depth of 24 m. CERN is located in the Pre Geneva basin, which is filled by sedimentary deposits called molasse. The chemical analysis of the molasse rock was carried out by the EMPA laboratory in Dübendorf (Switzerland) via X-ray fluorescence spectrometry (WD-XRF). Soil samples were dried before the XRF measurements. The XRF analysis is not reliable in quantifying the 1st and 2nd period elements (from H to Ne) in inorganic solid matrices (EMPA, 2011). For this reason, the oxygen and carbon contents were extrapolated using typical values of molasse soil (Vincke and Stevenson, 2000). The results of this chemical assay are shown in Table 1. The knowledge of the natural water content of the soil (moisture) is essential to discriminate the fraction of radioactivity leaching into groundwater from the soil, from the fraction coming from the moisture. The moisture was measured in CERN Environmental laboratory by drying a known amount of earth. The soil moisture content may be expressed by weight as the ratio of the mass of water present to the dry weight of the soil sample. To determine this ratio, the samples and container were weighed in the laboratory both before and after drying, the difference being the mass of water originally in the sample. The water content measured in the soil specimen was 5% by weight. Three cylindrical plastic containers 5 cm long and 4.5 cm in diameter were filled with 100 g of soil. A sample of drinking water was collected in a 50 ml plastic container 10 cm long and 2.5 cm in diameter (Fig. 1). The natural radioactivity in the samples was measured before irradiating them in the experiment. A sample of molasse was measured with a Ge detector for 5 104 s (about 14 h). The main contribution of the natural radioactivity comes from 40K (0.32 Bq g1) with small contributions from the thorium (0.018 Bq g1) and uranium chains (0.019 Bq g1). These results are in good agreement with the activity concentrations given in Vincke and Stevenson (2000). Additional information on the samples and on the chemical analysis can be found in La Torre et al. (2012a).
Table 1 Chemical composition of dried soil sample (density: ~1.4 g cm3). Element
O
Si
Ca
Al
C
Fe
Mg
K
Na
(g/100 g)
38.8a
24
16
6.8
5a
4
2
1.9
0.7
0.42
Mn
Ba
P
Sr
Zn
Cr
Zr
Eu
Ni
S
0.11
0.06
0.06
0.05
0.03
0.02
0.02
0.01
0.01
0.01
(g/100 g) a
Fig. 1. Soil and water samples used in the activation experiment.
Extrapolated value, not quantifiable by XRF analysis.
Ti
The activation experiment was carried out at the H4IRRAD facility in the North Experimental Area of the CERN Super Proton Synchrotron (SPS). This facility hosts a copper target struck by the SPS primary proton beam with momentum of 400 GeV/c and average intensity of 3 109 protons per pulse (over a supercycle of about 45 s and an extraction length of ~5 s). The soil and water containers were installed under the copper target. A schematic view of the experimental set up is shown in Fig. 2. The activation experiment was carried out at the H4IRRAD facility (Biskup et al., 2011) in CERN North Experimental Area. The soil and water containers were installed under the copper target struck by the SPS primary proton beam with momentum of 400 GeV/c and average intensity of 3 109 protons per pulse (over a supercycle of about 45 s and an extraction length of ~5 s). An argon ionisation chamber (XION) placed in the H4 beam line just upstream of the copper target monitored the intensity of the primary beam. One XIONcount corresponds to (6900 ± 690) particles impinging on the target (La Torre et al., 2012b). 3.1. Analysis of the irradiated samples Immediately after irradiation with ~ 7.5 1013 accumulated protons, the dose rate of the samples was of the order of a few mSv/ h. Most of this radioactivity was due to very short half-life radioisotopes. Since the radioisotopes of interest to this study have medium and long half-life, the samples were let decaying for 10 days before counting. The earth and water samples were measured with a high sensitivity, low-background, high-purity germanium (HPGe) detector by Canberra. The data acquisition and analysis was carried out using Canberra's Genie-2000 spectrometry software and the PROcount-2000 counting procedure software (Genie-2000 Software, 2000). This is a comprehensive software package for data acquisition, display and analysis, which includes a set of advanced spectrum analysis algorithms providing a complete analysis of gamma ray spectra. Three gamma spectrometry analyses were performed for each sample, at cooling times of one week, one month and two months. The soil samples could not be directly counted for 3H due to the low beta-particle endpoint energy (18 keV), which is absorbed in the sample. For this reason, the tritium activity in the soil was estimated via Monte Carlo calculations (Section 3.2). The tritium activity in the water was determined using a liquid scintillation counter (Packard TRI-CARB 3180TR/SL), measuring a mixture of 8 ml of activated water and 12 ml of so-called liquid scintillation cocktail (Packard Ultima Gold LLF). In case of high precision
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
9
measurements, distillation is usually recommended requiring well controlled conditions where other radionuclides present in the sample (e.g. 22Na) may significantly increase the result for tritium. This was not needed in the present case, as the potential interference of other radionuclides in the tritium pulse-height window was negligible. Tables 2 and 3 show the results of the gamma spectrometry and tritium measurements for the soil and water samples after 10 days of cooling time. Only radionuclides with the smallest experimental uncertainty and comparable cooling time and half-life are listed. Quoted errors include statistical and non-statistical uncertainties of the gamma spectrometry analysis. 3.2. FLUKA simulations The specific activities of the radionuclides in the samples were predicted by Monte Carlo simulations with the FLUKA code (Ferrari et al., 2005; Battistoni et al., 2007). The simulations were performed using a detailed geometrical model of the experimental facility implemented by the H4IRRAD team (Biskup et al., 2011). Fig. 2 shows the target station as implemented in the FLUKA simulations. The irradiated samples were reproduced in details with the actual size, the Plexiglas container and their position near the target. The elemental compositions of the samples were those obtained from the chemical analysis, to which 5% of water (moisture) was added. The full hadronic cascade was simulated in the irradiation area. The nuclear spallation process is the main reaction mechanism at 400 GeV energy. Tritium is produced by spallation from oxygen atoms in water and soil. The production of 22Na is due to spallation reactions from 23Na and heavier elements in soil (e.g. 24Mg, 27Al, 28 Si). For an accurate description of all nuclear processes relevant for isotope production in the soil and water samples, the coalescence mechanism and the evaporation of heavy fragments were explicitly turned on via two separated PHYSICS cards. The
Fig. 3. Fluence spectra of protons, neutrons and charged pions at the location of the soil and water samples.
coalescence mechanism describes the emission of energetic light fragments (up to alpha particles), while the production of fragments up to mass 24 is described by the evaporation of heavy fragments. The algorithm adopts a sampling scheme for the emitted particle spectra including sub-barrier effects and takes the full energy dependence of the nuclear level densities into account. These physics models allow a much more accurate description of the production of residual nuclei (Ferrari et al., 2005). To obtain an accurate simulation of the physical process involved in the activation experiment, the card DEFAULTS was used. The particle transport threshold was set at 100 keV, except for neutrons transported down to thermal energies. Low energy neutrons (below 20 MeV) were transported using the 260 group library. A total of 3 105 primary particle histories were simulated in 100 Monte Carlo runs. The total irradiation time was 1.73 106 s with an average beam intensity of 4.33 107 protons per second, i.e. about 7.5 1013 accumulated number of protons. Region-importance biasing was used in order to enhance the statistical accuracy of the results. The production of residual nuclei and their radioactive decay were scored in the same run. The RESNUCLEi card scores the residual nuclei produced in inelastic interactions, while the radioactive decay was calculated using the RADDECAY, DCYSCORE, DCYTIMES and IRRPROFIle cards, taking into account the decay chains, build-up of isotopes and the irradiation profile. The USRTRACK card scores the particle spectral fluences in the sample region. Fig. 3 shows the differential distributions of the energy fluence of neutrons, protons and charged pions. Experimental values, FLUKA predictions and their ratios are shown in Table 2 for the soil and in Table 3 for the water sample.
Table 2 Results of experiment and calculations for the specific activity in the soil sample. Nuclide
3
H Be Na 46 Sc 48 V 51 Cr 52 Mn 54 Mn 56 Co 58 Co 7
22
Fig. 2. Irradiation set-up in the FLUKA geometry drawn with SimpleGeo (Theis et al., 2006). The aluminium target holder is shown in green, the copper target in blue and the irradiated soil sample in brown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
t1/2
12.32 y 53.1 d 2.6 y 83.8 d 16 d 27.7 d 5.59 d 312 d 77.3 d 70.9 d
Specific activity (Bq g1)
Ratio
Experimental
FLUKA
Exp/FLUKA
e 425 ± 32 12.5 ± 0.8 3.02 ± 0.19 9.33 ± 0.50 25.0 ± 2.3 6.01 ± 0.43 6.02 ± 0.42 0.40 ± 0.04 0.22 ± 0.05
14.1 257 10.2 2.99 12.3 27.2 6.95 5.74 0.86 0.35
± ± ± ± ± ± ± ± ± ±
0.3 4 0.2 0.35 1.1 1.4 0.49 0.23 0.09 0.05
e 1.65 1.22 1.01 0.76 0.92 0.86 1.05 0.46 0.63
10
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
Table 3 Results of experiment and calculations for the specific activity in the water sample. Nuclide
3 7
H Be
t1/2
12.32 y 53.1 d
Specific activity (Bq g1)
Ratio
Experimental
FLUKA
Exp/FLUKA
28.9 ± 2.6 523 ± 48
27.6 ± 0.5 480 ± 12
1.05 1.09
4. Leaching procedures Radionuclides affecting the groundwater come from leaching. Two leaching possibilities were investigated (Fig. 4): 1) water stagnation with irradiated soil (mixing system); 2) water percolation through the irradiated soil (flowing system). In this work, we used the standard leaching procedure for the measurements of the leaching factor, consisting of 10 parts by weight of water and one part of irradiated soil (Borak et al., 1972; Baker, 1975, 1985; Sullivan, 1987; Malensek et al., 1993). To measure the amount of radioactivity passed from the irradiated soil to the water in both systems, leached water samples were systematically measured by gamma spectrometry and by scintillation analysis. 4.1. Soilewater mixing system In the mixing system, 100 g of the irradiated soil were placed in a graduated flask together with 1 l of distilled (tritium free) water (Fig. 4 e left). After vigorous shaking to disperse the soil in the water, the mixture was stirred for 8 h. To measure the radioactivity leached out, a 100 ml sample of the hazy water was filtered through a Millipore filter (0.45 mm). The gamma activity in the water was measured with a Germanium detector while b emitters were measured with a liquid scintillation counter after distillation.1 Activated soil and water were in contact for 4 months and the analyses on the leached water were repeated after 1, 2 and 4 months. 4.2. Soilewater flowing system In the flowing system (Fig. 4 e right) a sample of 100 g of irradiated soil was placed in a funnel with a Millipore filter (0.45 mm)
Table 4 Radioactivity measured in the leached water after correction for the decay and the concentration for the soilewater mixing system. Mixing time
Specific activity (Bq l1)
8 1 2 4
518 541 542 542
3 H (t1/2 ¼ 12.32 y)
h month months months
± ± ± ±
31 32 32 33
7
Be (t1/2 ¼ 53.1 d)
22
5.84 ± 3.03
96.7 131 136 144
Na (t1/2 ¼ 2.6 y) ± ± ± ±
5.8 8 8 9
48 V (t1/2 ¼ 16 d)
8.21 ± 0.49 10.2 ± 1.3 16.1 ± 2.1
MDA: minimum detectable activity.
connected to a graduated container. A glass separatory funnel was placed over the funnel and filled with 1 l of distilled (tritium free) water. The glass stopcock allowed controlling the rate of addition of the water to 125 ml h1. The gamma activity in the water was measured with a Germanium detector while b emitters were measured with a liquid scintillation counter after distillation. The whole procedure was repeated after 1, 2 and 4 months. 5. Data analysis The experimental results were analysed and corrected for the decay time and for the radioactivity concentration. The error bars shown in Tables 4 and 5 were estimated using the propagation of uncertainty law based on the uncertainties of the results of the gamma spectrometry and liquid scintillation measurements that include statistical and non-statistical errors. The specific activities measured in the water as a function of the mixing time for the soilewater mixing system are given in Fig. 5 and in Table 4. The 3H concentration in the water quickly increases to 95.6% (518 Bq l1) of the total activity after 8 h of mixing time. It takes one month to rise to 99.8% (541 Bq l1) and 2 months to reach 542 Bq l1, the maximum value measured in the water. A similar behaviour was found for 22Na. The highest concentration was measured after 123 days with 144 Bq l1. Eight hours are enough to reach 67% (96.7 Bq l1) of the total 22Na activity measured in the leached water. Half of the total 48V activity was measured after 8 h and the maximum activity is measured after 61 days with 16.1 Bq l1. The results of the water analysis for the soilewater flowing system are given in Fig. 6 and in Table 5. The specific activity of 3H decreases from 384 Bq l1 to 3.23 Bq l1 after the first wash. The same behaviour is observed for 22Na that passes from 126 Bq l1 to 12.7 Bq l1 in one wash. The concentration of 48V and 51Cr varies little even after three washes: 48V slowly decreases from 4.51 Bq l1 to 1.95 Bq l1 whereas 51Cr decreases from 13.0 Bq l1 to 8.33 Bq l1. 5.1. Fraction of radioactivity leached into the water The activity concentration of the leached water for both systems was compared with the radioactivity measured in the activated soil Table 5 Radioactivity measured in the leached water including the decay correction for the soilewater flowing system. Wash Specific activity (Bq l1) 3 7 22 48 51 H Be Na V Cr (t1/2 ¼ 12.32 y) (t1/2 ¼ 53.1 d) (t1/2 ¼ 2.6 y) (t1/2 ¼ 16 d) (t1/2 ¼ 27.7 d)
Fig. 4. Soilewater mixing system (left) and flowing system (right).
1 100 ml of leached water were evaporated to dryness. During this process approximately 20 ml of condensate was collected for 3H measurement.
1 2 3 4
384 3.23 1.67 2.05
± ± ± ±
23 0.93 0.90 1.09
8.48 ± 2.40
MDA: minimum detectable activity.
126 12.7 3.15 4.71
± ± ± ±
6 0.9 0.22 0.54
4.51 ± 0.28 2.55 ± 1.4 1.95 ± 1.23
13.0 ± 3.2 9.16 ± 4.57 8.33 ± 3.67
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13
11
Table 6 Fraction of radioactivity leached by water in the mixing system. Nuclide t1/2
tot Leachable fraction for mixing (%) ½Atot i ðleached waterÞ=Ai ðsoilÞ
8h 3
a
H Be Na 48 V 7
22
12.32 y 36.7 ± 7.7 53.1 d 0.014 ± 0.007 2.6 y 7.74 ± 0.68 16 d 0.88 ± 0.08
1 month
2 months
4 months
38.4 ± 8.0 e 10.5 ± 0.9 1.09 ± 60.16
38.5 ± 8.1 e 10.9 ± 1.0 1.73 ± 0.44
38.5 ± 8.1 e 11.5 ± 1.1 e
a 3 H activity measured in the leached water compared with the activity estimated with FLUKA in soil.
the tritium activity measured in the activated water sample (Table 3). Since the measured soil moisture is 5% by weight (see Section 2), i.e. 5 g for 100 g of soil, it can be assumed that:
Atot trit ðmoistureÞ ¼ Atrit ðwaterÞ 5 g ¼ ð145±9Þ Bq Fig. 5. Specific activity in the leached water as a function of the mixing time for the soilewater mixing system, after correction for the decay and the concentration.
in order to estimate the fraction leached out. Tables 6 and 7 show the ratio between the total activity measured in the leached water and the total activity measured in the activated soil for each tot radionuclide ½Atot i ðleached waterÞ=Ai ðsoilÞ for the mixing and flowing systems, respectively. Since the liquid scintillation analysis cannot be performed on the soil, the tritium activity measured in the leached water was compared with the tritium estimated in the soil by the FLUKA simulations. Fig. 7 shows the leachable fraction of 3 H and 22Na, the two radionuclides of major interest in this study. Most of the radioactivity leached into the water just after 8 h of mixing time (mixing system) or after the first wash (flowing system). In the first case, the longer the soil is mixed with the water, the more radioactivity is leached out. After two months of stirring, 39% of the 3H and 11% of the 22Na is leached by the water. For the flowing system, the leachable fraction after one wash is 27% for 3H and 10% for 22Na. As from the second wash, the leachable fraction falls down to a fraction of per cent for both radioisotopes. It is clear that the mixing system is a much more efficient way to extract radioactivity from the activated soil. It is also interesting to compare the tritium extracted by water with its activity in the bulk (dried soil) and in the moisture of the activated soil. The tritium activity in the soil moisture Atot ðmoistureÞ can be estimated from trit
(1)
where Atrit(water) is the specific activity of 3H in the activated water, i.e. (28.9 ± 2.9) Bq g1. The total activity of 3H in the soil, Atot ðsoilÞ, is given by: trit
Atot trit ðsoilÞ ¼ Atrit ðsoilÞ 100 g ¼ ð1410±30Þ Bq
(2)
where Atrit(soil) is the specific activity of 3H estimated by FLUKA in the soil (Table 2), i.e. (14.1 ± 0.3) Bq g1. The tritium activity in the bulk of the activated soil Atot trit ðbulkÞ can thus be estimated as follow: tot tot Atot trit ðbulkÞ ¼ Atrit ðsoilÞ Atrit ðmoistureÞ ¼ ð1265±105Þ Bq
(3)
where Atot ðmoistureÞ and Atot ðsoilÞ are given by expressions (1) trit trit and (2). Tritium produced in soil bulk is much more difficult to extract than tritiated water moisture, which behaves chemically like water (Canadian Nuclear Safety Commission (CNSC), 2013). For this reason, it can be assumed that both techniques extracted before the tritium produced in the soil moisture so that the remaining part of tritium can be attributed to the soil bulk. Table 8 summarizes the results and shows the leachable fraction of the moisture and the leachable fraction of the bulk for the mixing and flowing systems. According to the previous assumption, it could be inferred that both systems leached out 100% of tritium produced in the moisture; the mixing system removed 32% of tritium produced in the bulk, whereas the flowing system removed only 20%. 6. Comparison with literature data Borak et al. (1972) report a series of experiments carried out to determine the most practical approach for measuring the leachability of elements from the soil (glacial till). Finally, a batch processing technique by directly mixing 10 parts by weight of water and one part of soil was used. Baker et al. (1997) tested several soil Table 7 Fraction of radioactivity leached by water in the flowing system. Nuclide
t1/2
Leachable fraction for flowing (%) tot ½Atot i ðleached waterÞ=Ai ðsoilÞ
3
12.32 y 53.1 d 2.6 y 16 d 27.7 d
27.2 0.02 10.1 0.48 0.52
1 wash Ha 7 Be 22 Na 48 V 51 Cr Fig. 6. Specific activity in the leached water as a function of wash including decay time correction for the wateresoil flowing system.
± ± ± ± ±
5.6 0.01 0.8 0.04 0.13
2 washes 0.23 e 1.02 0.27 0.37
± 0.08 ± 0.10 ± 0.14 ± 0.18
3 washes 0.12 e 0.25 0.21 0.33
± 0.07 ± 0.02 ± 0.13 ± 0.15
4 washes 0.15 ± 0.08 e 0.38 ± 0.04 e e
a 3 H activity measured in the leached water compared with the activity estimated with FLUKA in soil.
12
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13 Table 9 Comparison of the present results for 3H and
3
H
22
Na
22
Na with literature data.
This work (mixing)
This work (flowing)
Borak et al. (1972)
Baker et al. (1997)
39% (100% moisture, 31% bulk) 12%
28% (100% moisture, 20% bulk) 11%
100%
66e100%
10e20%
7e32%
components. This study was carried out for a specific type of soil, the molasse, and the results of leaching of radioactivity from the soil bulk should not be generalized to other type of soils without a dedicated investigation. 7. Conclusions
Fig. 7. Cumulative fraction of 3H and vated soil.
22
Na activities leached by water from the acti-
samples (Weathered Austin chalk, Eagle Ford shale, Ellis County soil) using the same technique of Borak et al., but mixing an amount of distilled water equal in weight to the weight of the sample. Both authors determined the amount of 3H produced in the soil by distillation of the activated soil in an oven and collecting the water driven off. This water was then assayed for 3H by means of liquid scintillation counting. Borak et al. found that the leached water accounted for all of the 3 H measured by the bake out process, on the other hand the fraction of 22Na removed from the activated soil was about 10e20%. Baker et al. measured a 3H percentage leachable in the range of 66e100% and an amount of 22Na in the leached water ranging between 7% and 32%. Table 9 compares the present results with those of Borak et al. and Baker et al. for 3H and 22Na. If the 22Na measured in the leached water is compatible with the literature data, a large discrepancy is observed for 3H. Borak et al. and Baker et al. made the assumption that all the 3H produced in the sample that could leach was transferred to the water that comes out in the distillation process. This assumption leads to an overestimation of the total 3H leachable fraction, since the evidence suggests that not all the tritium produced from soil molecules can be extracted with the distillation process. This work seems to show that 100% of the 3H produced in the moisture is measured in the leached water, but only 31% (mixing) and 20% (flowing) of 3H produced in soil bulk can leach out. It is important to mention that other parameters can influence the leaching process of radioactivity from the soil. For example, the soil composition and particularly its organic fraction can have a significant impact on the retention of dissolved radioactive
Table 8 Tritium percentage leachable from soil bulk and moisture for both systems. Activity (Bq) 3
Soil bulk (95 g) Soil moisture (5 g) Leached water Fraction of 3H moisture leached out Fraction of 3H bulk leached out
H (mixing)
1265 ± 105 145 ± 9 542 ± 33 100% 31.4%
3
H (flowing)
384 ± 23 100% 18.9%
CERN soil and water samples were irradiated in the H4IRRAD test facility. Gamma spectrometry and beta scintillation analysis were performed to determine the radioactivity induced in the samples. The main radionuclides found in the soil after 10 days of cooling time are 7Be, 22Na, 46Sc, 48V, 51Cr, 52Mn, 54Mn. The water radioactivity is due to 3H and 7Be radionuclides. FLUKA simulations were benchmarked against the experimental results and allowed estimating the concentration of 3H in the soil. Two leaching procedures were used and compared to quantify the amount of radioactivity leached out of the soil into the water. The mixing system was able to remove up to 39% of 3H and 12% of 22Na from the irradiated soil. The flowing system extracted 28% of 3H and 11% of 22 Na. The results seem to indicate that 100% of the tritium produced in the soil moisture may leach out, but only 31% (mixing) and 20% (flowing) of the tritium produced in the soil bulk can be extracted. Thus most of the transferable 3H produced in soil depends on the amount of water in the soil at the time of irradiation. On the other hand, only a small percentage (31% in mixing and 20% in flowing) of tritium, which is soil-type dependent, can leach out. This work provides previously unavailable information on the tritium leachable fraction from activated soil. Acknowledgements We wish to thank P. Vojtyla for providing access to the Environmental Laboratory and for many helpful discussions. We would like to thank A. Dziewa and F. Malacrida for performing the gamma spectrometry and the liquid scintillation measurements. We are grateful to M. Calviani, M. Brugger and the H4IRRAD team for allowing the irradiation of the samples and for providing the FLUKA geometry of the facility. References Agosteo, S., Magistris, M., Silari, M., 2005. Radiological considerations on multi-MW targets, part I: induced radioactivity. Nucl. Instrum. Methods Phys. Res. A 545, 813e822. Baker, S.I., 1975. Soil Activation Measurements at Fermi Laboratory (Internal report). Fermi National Accelerator Laboratory. Baker, S.I., 1985. Fermilab soil activation experience. In: Paper Presented at the Fifth DOE Environmental Protection Information Meeting, Albuquerque, New Mexico, November 1984, CONF-841187, pp. 673e683. Baker, S., Bull, J., Gross, D., 1997. Leaching of accelerator produced radionuclides. Health Phys. 73, 912e918. Battistoni, G., Muraro, S., Sala, P.R., Cerutti, F., Ferrari, A., Roesler, S., Fasso, A., Ranft, J., 2007. The FLUKA code: description and benchmarking. In: Albrow, M., Raja, R. (Eds.), Proceedings of the Hadronic Shower Simulation Workshop 2006, Fermilab, 6e8 September 2006, AIP Conference Proceeding, vol. 896, pp. 31e49. Biskup, B., Brugger, M., Calviani, M., Efthymiopoulos, I., Kwee, R., Mekki, J., La Torre, F.P., Lebbos, E., Mala, P., Manessi, G., Nordt, A., Pozzi, F., Roeed, K., Severino, C., Silari, M., Thornton, A., 2011. Commissioning and Operation of the H4IRRAD Mixed-field Test Area (CERN technical note, CERN-ATS-Note-2011-121 PERF).
F.P. La Torre, M. Silari / Journal of Environmental Radioactivity 143 (2015) 7e13 Borak, T.B., Awschalom, M., Fairman, W., Iwami, F., Sedlet, J., 1972. The underground migration of radionuclides produced in soil near high-energy protons accelerators. Health Phys. 23, 679e687. Canadian Nuclear Safety Commission (CNSC), 2013. Environmental Fate of Tritium in Soil and Vegetation, ISBN 978-1-100-22687-3. EMPA e Swiss Federal Laboratories for Material Science and Technology, 2011. Laboratory for Solid State Chemistry and Catalysis, Überlandstrasse 129, CH8600 Dübendorf, Test Report n. 459595. Ferrari, A., Sala, P.R., Fasso, A., Ranft, J., 2005. FLUKA: a Multi-particle Transport Code (CERN-2005-10, INFN/TC_05/11, SLAC-R-773). Genie-2000 Spectroscopy Software. ISO 9001. Canberra Industries. http://www. canberra.com. Hoyer, F., 1968. Induced Radioactivity in the Earth Shielding on Top of High-energy Particle Accelerators (CERN 68-42). La Torre, F.P., Magistris, M., Silari, M., 2012a. Irradiation of Soil Samples at H4IRRAD to Study Radioisotope Leaching into Groundwater: Preliminary Study (CERN technical report, CERN-DGS-2012-047-RP-TN). La Torre, F.P., Manessi, G.P., Pozzi, F., Severino, C.T., Silari, M., 2012b. Cu and Al Activation Experiments for Beam Monitoring in H4IRRAD (CERN technical report, CERN-DGS-2012-022-RP-TN). Malensek, A.J., Wehmann, A.A., Elwyn, A.J., Moss, K.J., Kesich, P.M., 1993. Groundwater Migration of Radionuclides at Fermilab (Fermilab report, FERMILAB-TM1851). Nelson, W.R., 1965. Radioactive Groundwater Produced in the Vicinity of Beam Dumps (Stanford linear accelerator, center international note SLAC-TN-65-16). Racky, B., Dinter, H., Leuschner, A., Tesch, K., 1998. Radiation environment of the linear collider TESLA. In: Proceedings of the Fourth Workshop on Simulating Accelerator Radiation Environments (SARE-4), Knoxville, Tennessee, U.S.A., p. 14. Ranft, J., Goebel, K., 1970. Estimation of Induced Radioactivity Around High-energy Accelerators from Hadronic Cascade Star Densities Obtained from Monte Carlo Calculations (CERN HP-70-92).
13
Rindi, A., 1972. Induced Radioactivity in the Cooling Waters of the CERN 300 GeV SPS (CERN international report Lab II-RA/Note/72-71). Rokni, S., Liu, J.C., Roesler, S., May 2000. Initial Estimates of the Activation Concentration of the Soil and Groundwater Around the NLC Beam Delivery System Tunnel (SLAC RP note RP-00-04). Stapleton, G.B., Thomas, R.H., 1972. Estimation of the induced radioactivity of the groundwater system in the neighbourhood of a proposed 300 GeV high-energy accelerator situated on a chalk site. Health Phys. 23, 689. Sullivan, A.H., 1987. Groundwater Activation Around the AA and ACOL Target Areas (CERN internal report CERN/TIS-RP/IR/87-34). Tesch, K., 1997. Production of Radioactive Nuclides in Soil and Groundwater Near Dump of a Linear Collider (DESY internal report-DESY, D3-D86). Theis, C., Buchegger, K.H., Brugger, M., Forkel-Wirth, D., Roesler, S., Vincke, H., 2006. Interactive three dimensional visualization and creation of geometries for Monte Carlo calculations. Nucl. Instrum. Methods Phys. Res. A 562, 827e829. Thomas, R.H., 1970. Possible Contamination of Groundwater System by High-energy Acceleration (Internal report UCRL-20131). Lawrence Berkley Laboratory. United States General Accounting Office, 1997. Information on the Tritium Leak and Contractor Dismissal at the Brookhaven National Laboratory (Department of Energy, GAO/RCED-98-26). Vincke, H., Stevenson, G.R., 2000. Production of radioactive isotopes in molasse. In: 5th Meeting of the Task Force on Shielding Aspects of Accelerators, Targets and Irradiation Facilities, Paris, France, 18e21 Jul 2000, pp. 47e57. Vogt, K., Haida, M., Fehrenbacher, G., 2008. Soil activation studies for the FAIR project. In: 1st ARIA 2008 Workshop on Accelerator Radiation Induced Activation, October 13e17, 2008. Paul Scherrer Institut (PSI), Switzerland. Warren, G.J., Busick, D.D., McCall, R.C., 1969. Radioactivity produced and released from water at high energies. In: Proc. Second Int. Conf. in Accelerator Dosimetry, Stanford, November 1969, CONF-691101. Wehmann, A., Childress, S., 1999. Tritium Production in the Dolomitic Rock Adjacent to NuMI Beam Tunnels (Fermilab report, FERMILAB-TM-2083).