1 INTRODUCTION
Tritium occurs in the environment as tritiated water (HTO), organically bound tritium (OBT), etc. Tritium is easily absorbed by the human body through ingestion, inhalation and skin absorption [1]. The emitted β-particles can cause radiation damage to the human body through internal exposure [2]. Hence, tritium release is one of the most concerning topics in nuclear power plants.
The molten salt reactor (MSR) is a new class of nuclear fission reactor that uses a fluid molten salt mixture as fuel. It is one of the six advanced reactor types for future nuclear energy proposed in the Generation IV International Forum (GIF). The MSR design has aroused widespread concern recently [3]. Indeed, the Chinese Academy of Sciences launched the thorium molten salt reactor nuclear energy system (TMSR) research program in 2011, which includes solid-fueled MSR (MSR-LF) and liquid-fuel MSR (MSR-LF) [4,5]. A 2 MW liquid-fueled molten salt experimental reactor (TMSR-LF1) is the goal of the MSR-LF in the start-up stage. In TMSR-LF1, the molten salt of LiF-BeF2-ThF4-UF4 is used as the fuel. Here, the abundance of 7Li is 99.95%. A certain amount of tritium is generated during normal operating conditions. This may be released into the environment because of its permeability. It is also necessary to analyze the tritium production in TMSR-LF1 as well as its environmental impacts.
Here, we calculated the tritium production in TMSR-LF1 by running ORIGEN-S [6] modules with an updated cross section library generated in TRITON [7] of the SCALE 6.1.3[8] code system. Tritium production was compared to other reactors. Its distribution in the atmosphere and exposure was estimated under normal operation. This study describes the amount of tritium radiation released from the TMSR-LF1 to the public and provides useful information on tritium for the TMSR program.
2 DESCRIPTION OF TMSR-LF1
As an optimized MSR based on the molten salt reactor experiment (MSRE) [9,10], the TMSR-LF1 is a graphite-moderated reactor that is composed of a core and a reflector. The thermal power is 2 MW, and the fuel salt is LiF-BeF2-ThF4-UF4 (68-28-0.1-3.9 mole%) with 99.95% abundance of 7Li and a coolant salt of LiF-NaF-KF. Table 1 lists the principal physical parameters of the TMSR-LF1.
| Parameters | Data |
|---|---|
| Thermal power | 2 MW |
| Average temperature | 610°C |
| 235U enrichment | 16% |
| 235U loading | 39.6 kg |
| Cross sectional area of fuel salt passage | 2.385 cm2 |
| Fuel salt volume at active zone | 1.84×105 cm3 |
| Active zone diameter | 110 cm |
| Active zone height | 110 cm |
| Reactor vessel diameter | 195 cm |
| Reactor vessel height | 250.6 cm |
The active zone is a 110-cm-diameter by 110-cm-high graphite matrix structure made up of 368 graphite stringers. The graphite stringers are 5.1 cm by 5.1 cm in cross section and about 110-cm-long with half-channels consisting of two 1/2 cylinders and a cuboid machined in the four faces of each stringer to form flow passages (Fig. 1). There are 696 fuel salt passages in the core. There is an upper plenum and a lower plenum above and below the active zones, respectively. The graphite reflector surrounds the active zone, and all metals contacted with the molten salt are Hastelloy-N including the reactor container, structure materials, etc. The control rods, neutron source channel and measurement channel are located on the graphite reflector (Fig. 2).
3 TRITIUM PRODUCTION
3.1 Method of analysis
In TMSR-LF1, the tritium is produced in the reactor core through the reaction of 6Li (n, α) T, 7Li (n, n’, α) T, and 19F (n, n’,T) 17O, ternary fission of 235U, 238U, and 232Th as well as the direct and indirect 9Be source. The tritium production of TMSR-LF1 was calculated via ORIGEN-S modules with an updated cross section library generated by TRITON [11] that couples the neutron transport and depletion in the SCALE 6.1.3 code system.
The time of continuous full power operation was 180 days according to the design scheme. The ENDF/B-VII.1 238-group library was adopted to carry out neutron transport calculations, and the ORIGEN-S data libraries were used to analyze the burn-up. One-group cross sections were processed via CENTRM, and the neutron spectum averaged over the fuel salt in the core were updated during operation (Fig. 3).
To verify the reliability of the method, the tritium yield of MSRE was calculated with 8 MW power with the fuel salt of LiF-BeF2-ZrF4-UF4 (65-29.2-5-0.8 mole%) and 0.0074% abundance of 6Li[12,13]. The results are shown in Table 2.
| Source | Literature data[12] | Calculation data |
|---|---|---|
| 6Li | 32 | 30 |
| 7Li | 5 | 5 |
Table 2 shows that the calculation result of the 7Li tritium production rate is in reasonable agreement with the literature. There is little difference in the 6Li tritium production rate. The deviation value between the literature data and the calculated data is 6%—this may be because of different library cross-sections or the variable calculation precision values from different software. In the literature, the cross section of 6Li(n, α)T reaction is 476 barns for the thermal neutron energy spectrum and 475 barns in the SCALE6.1.3 program.
3.2 Tritium production and its distribution
The one-group cross sections of tritium-generating reactions at the beginning and the end of the cycle are listed in Table 3. The tritium production generated from each reaction is shown in Fig. 4.
| Reaction | One-group cross section | |
|---|---|---|
| 0 day | 180 day | |
| 6Li(n, α) T | 6.621×101 | 6.601×101 |
| 7Li(n, n’, α)T | 2.009 ×10-3 | 1.967×10-3 |
| 19F(n, n’, T)16O | 2.212×10-6 | 2.670×10-6 |
| 9Be(n, T)7Li | 5.013×10-7 | 6.113×10-7 |
| 9Be(n, α)6He | 4.440×10-3 | 4.435×10-3 |
| 235U ternary fission(T emission) | 3.970×10-3 | 3.958×10-3 |
| 238U ternary fission(T emission) | 4.104×10-6 | 4.102×10-6 |
| 232Th ternary fission(T emission) | 7.174×10-7 | 7.157×10-7 |
The reaction of 6Li(n, α)T is characterized by a large cross section—66.21 barns at the beginning of the cycle. Thus, a large amount of tritium is produced, though there are few 6Li atoms in the molten salt. Furthermore, the 6Li is replenished via the β decay of 6He that is produced through the reaction of 9Be(n, α)6He. At the same time, the 9Be directly generates tritium through the (n, T) reaction. The one group cross sections of 7Li(n, n’, α)T and tritium-generating 235U ternary fission are on the same order of magnitude. Because of abundant 7Li atoms in the fuel salt, there are many neutron absorption events that produce a considerable amount of tritium. The reaction of 19F(n, n’, T)16O also produces tritium, but the rate of tritium formation is very low due to a much smaller cross section than 6Li.
Tritium production increases nearly linearly with the full power running time. The tritium production rate is about 8.90×1011Bq/d at the end of cycle. By calculating the contribution of each reaction at the end of the cycle, we found that the tritium generated from the reaction of 6Li (n, α)T is about 94% of the total. The others include: 5.7% from the reaction of 7Li (n, n’, α)T; 0.11% from 235U, 238U, and 232Th; 0.08% from 9Be; and 0.01% from 19F (n, n’,T)16O.
The literature suggests that tritium will be presented as molecules of tritium gas (T2) or tritium fluoride gas (TF) dissolved in the primary salt after generation [14]. For MSRE, the observed tritium amounted to 80% of the tritium production rate. There was 48% discharge from the fuel off of the gas system, 2% discharge from the coolant off gas system, 7% discharge from the coolant radiator air, 9% in the cell atmosphere, and 14% going into the core graphite [15]. Tritium in the off-gas system will be processed, while the tritium diffused through pipes or containment vessels may be released into the environment. The tritium distribution in TMSR-LF1 refers to that from MSRE because of its similar structure.
Therefore, we conservatively estimated that 36% tritium (including 20% unobserved tritium, 7% tritium in the coolant radiator air, and 9% in the cell atmosphere) was released into the environment during normal operations with a release rate of approximately 3.2×1011 Bq/d.
3.3 Comparison of tritium production with other reactors
Table 4 shows the tritium production of three types of reactors: MSRs (TMSR-LF1, MSRE and MSBR), PHWR (Qinshan Phase III) and PWR (Tianwan nuclear power plant). The normalized tritium production rate from TMSR-LF1 is the largest of these reactors, but its tritium production rate is the lowest because of the lowest thermal power. Of the MSRs, the normalized tritium production rate in TMSR-LF1 is 4.45×1011 Bq/(MW·d). It is 2.74×1011 Bq/(MW·d) in MSRE and 3.98×1010 Bq/(MW·d) in MSBR. The normalized tritium production rate of TMSR-LF1 is about 1.6 times that of MSRE and 11 times that of MSBR. The difference in tritium production among MSRs is mainly due to the different 6Li concentrations. Furthermore, 6Li (n, α)T is the main tritium-producing reaction. The abundance of 6Li is 0.05% in TMSR-LF1, and 0.0074%[13] in MSRE; it is 0.005%[16] in MSBR.
| TMSR-LF1 | MSRE [15] | MSBR [15] | PHWR [17] (Qinshan phase III) | PWR [18] (Tianwan) | |
|---|---|---|---|---|---|
| Thermal power (MW) | 2 | 7.3 | 2250 | 2158.5 | 3000 |
| Production rate (Bq/d)a | 8. 90×1011 | 2.00×1012 | 8.95×1013 | 1.80×1014 | 2.46×1012 |
| Normalized production rate (Bq/(MW·d)) | 4.45×1011 | 2.74×1011 | 3.98×1010 | 8.34×1010 | 8.20×108 |
4 ENVIRONMENTAL IMPACT ASSESSMENT
4.1 Tritium concentration in the atmosphere
Referring to the China experimental fast reactor (CEFR), the release height was set as 60m[19] to estimate tritium concentration with meteorological data over a defined area. The wind direction and stability classes over one year are shown in Fig. 5.
In the atmosphere, tritium undergoes downwind transport and turbulent diffusion. The plume model of the PC-CREAM 08 computer code [20] uses a standard Gaussian plume dispersion model and was applied to analyze the annual dispersion of tritium in the environment. The estimated annual average tritium concentration distribution is shown in Fig. 6. The maximum tritium concentration is 1.4 Bq/m3 in the western sector and 800 m from the reactor.
Tritium concentrations were also calculated under extreme conditions assuming only atmospheric stability class A. Stronger turbulence intensity results in a faster dispersion in the vertical direction. At the same time, low wind speed and calm was seen with stability class A. This weakens the transport in the downwind direction. These will form a high contamination area near the release source. Fig. 7 shows the ground-level concentration distribution of gaseous tritium under extreme conditions. The maximum value of the tritium concentration is 11.8 Bq/m3. This occurred 300 m west of the reactor. This is one order of magnitude higher than under normal conditions.
4.2 Dose assessment
The tritium released from TMSR-LF1 is mainly a gas. We assumed that all released T2 gas was converted into HTO during migration. The public dose of tritium from chronic atmospheric release mainly stems from internal exposure due to ingestion of contaminated food, drinking water, inhalation, and skin absorption from polluted air. A schematic diagram of tritium exposure pathways due to the nuclear facility is shown in Fig. 8. OBT is a chemical form of tritium and has a long biological half-life in the human body with a dose coefficient of ingestion in the human body that is about 2-fold that of HTO. Therefore, both forms (HTO, OBT) are considered here.
4.2.1 Inhalation and skin absorption
The internal exposure dose from inhalation to the public is given by equation (1) [21]:
where Din is the dose from inhalation, Sv/a; Cair is the concentration of HTO in air, Bq/m3; R is the inhalation rate, valued 0.96 m3/h; and DCFin is the adult inhalation dose coefficient of HTO set as 1.8×10-11 Sv/Bq [20]. The annual intake of HTO absorbed by skin is about 60% of the intake by inhalation [22].
4.2.2 Ingestion
Ingestion includes ingestion of contaminated plant products, animal products and drinking water. HTO could be transferred into OBT via photosynthesis and finally ingested by people. Hence, tritium is divided into HTO and OBT to estimate the ingestion dose. Dose from ingestion is according to formula (2) [21]:
where Ding is the ingestion dose, Sv/a; Qi, Qw, is annual consumption of food i, drinking water, respectively, kg/a;
| HTO | OBT | |
|---|---|---|
| Grains | 6.7 | 29.5 |
| Beans | -- | -- |
| Vegetable | 58.2 | 3.3 |
| Fruit | 48.7 | 5 |
| Meat | 9.4 | 8.5 |
| Milk | 20.3 | 1.6 |
| Egg | 13.8 | 4.1 |
| Consumption | |
|---|---|
| Grains | 164 |
| Beans | 9 |
| Vegetable | 131 |
| Fruit | 29 |
| Meat | 29 |
| Milk | 16 |
| Egg | 13 |
| Drinking water | 365 |
The effective doses due to the three exposure pathways (inhalation, skin absorption and drinking water) were calculated according to Eqs. (1) and (2). The calculated dose contribution from the exposure pathway of gaseous discharge to the general public is shown in Table 7.
| Exposure pathway | HTO | OBT | Total |
|---|---|---|---|
| Inhalation | 19% | -- | 19% |
| Skin absorption | 12% | -- | 12% |
| Ingestion food | 27% | 35% | 62% |
| Drinking water | 7% | -- | 7% |
| Total | 65% | 35% | 100% |
Table 7 shows that ingestion of food is the dominant exposure pathway and makes up 62% of the total effective dose; 19% of the total dose is from inhalation, 12% is from skin absorption, and 7% is from drinking water. Therefore, exposure is mainly from food intake due to the chronic atmospheric release of tritium. The data also indicate that 35% of OBT exposure comes from ingesting food. Thus, the OBT should not be ignored when estimating the impact of tritium. The maximum exposure doses to the public through two conditions are shown in Table 8.
| Inhalation | Skin absorption | food | Drink water | Total | |
|---|---|---|---|---|---|
| Normal condition | 0.19 | 0.12 | 0.63 | 0.07 | 1.01 |
| Extreme condition | 1.79 | 1.07 | 5.32 | 0.55 | 8.73 |
Table 8 shows that the maximum individual dose to the public stemming from tritium is about 1 μSv/a under normal conditions with a tritium concentration of 1.4 Bq/m3. This is about 9 μSv/a under extreme conditions with a tritium concentration of 11.8 Bq/m3.
5 CONCLUSION
Here, tritium production was calculated via ORIGEN-S by reading the cross section library generated by TRITON. This method was verified by comparing calculation data and literature data for tritium production in the MSRE. For TMSR-LF1, the calculated tritium production rate and normalized tritium production rates are 8.90×1011 Bq/d and 4.45×1011 Bq/(MW·d), respectively. The normalized tritium production of TMSR-LF1 is the largest among the compared MSRs due to its higher abundance of 6Li. Moreover, evaluation of the environmental impacts during normal operation is carried out based on the 36% release rate referring to MSRE. The maximum concentration of tritium is 1.4 Bq/m3 under normal condition, and 11.8 Bq/m3 under extreme conditions. The corresponding individual doses are about 1 μSv/a and 9 μSv/a, respectively. Meanwhile, the dose arising from food ingestion accounted for over 60% of total individual dose including 35% stemming from OBT. Thus, the ingestion of food is the main exposure pathway. OBT as a form of tritium should be not ignored when assessing chronic atmospheric release of tritium. This work provides a fundamental reference for radiation protection and environmental assessment of TMSR-LF1.
Tritium hazard report: pollution and radiation risk from Canadian nuclear facilities
.Development of a novel system for monitoring tritium in gaseous form
. Nucl. Sci. Tech. 26(2), (2015). doi:10.13538/j.1001-8042/nst.26.020602A technology roadmap for Generation IV nuclear energy systems
[C].Advanced fission energy program-TMSR nuclear energy system
[J], Bulletin of Chinese Academy of Sciences, 2012. 27(3): p. 366-374. DOI: 10.3969/j.issn.1000-3045.2012.03.016(in Chinese)Impact analysis of criticality safety for 10-MWt solid thorium-based molten salt reactor spent nuclear fuel storage system
. Nucl. Tech. 38(5), (2015). doi: 10.11889/j.0253-3219.2015.hjs.38.050602. (in Chinese)ORNL/TM-2005/39, ORIGEN-S: Depletion Module to Calculate Neutron Activation, Actinide Transmutation, Fission Product Generation, and Radiation Source Terms
[R].ORNL/TM-2005/39, TRITON: A Multipurpose Transport, Depletion, and Sensitivity and Uncertainty Analysis Module
[R].ORNL/TM-2005/39, Scale: A Comprehensive Modeling and Simulation Suite for Nuclear Safety Analysis and Design
[R].ORNL-TM-0728, MSRE design and operations report part I: Description of reactor design
[R].Simulation of radiation dose distribution and thermal analysis for the bulk shielding of an optimized molten salt reactor
. Nucl. Sci. Tech. 26(4), (2015). doi: 10.13538/j.1001-8042/nst.26.040603Issues in the calculation of the tritium production of the fluoride-salt-cooled high-temperature reactors using SCALE
. Nuclear Techniques, 2015, 38(8). DOI: 10.11889/j.0253-3219.2015.hjs.38.080601(in Chinese)ORNL-CF-70-2-7, Tritium in the MSRE: Calculated Production Rates and Observed Amounts
[R].ORNL-0730, MSRE design and operations report. Part III Nuclear analysis
[R].ORNL-4548, Molten-salt reactor program semiannual progress report
[R].ORNL-5759, Distribution and behavior of tritium in the coolant-salt technology facility
[R].ORNL-4541, Conceptual design study of a single-fluid molten-salt breeder reactor
[R].Analysis of effluence tritium release from Qinshan phase III heavy water reactor
[J]. Radiation Protection, 2009. 29: p. 80-85 (in Chinese).Estimate of annual yields of 14C and 3H produced in Tianwan NPP
[J]. Radiation Protection Bulletin, 2004. 24: p. 23-26 (in Chinese).Monitoring system design of the radioactive gas emitted from the CEFR
inRadilolgical environmental impact assessment for tritium discharged from inland NPPs
. Radiation Protection, 2012. 32(2): p. 66-69 (in Chinese).Tritium in the Canadian environment: levels and health effects. RSP-0153-1
. 2002 http://www.odwac.gov.on.ca/standards_review/tritium/Osborne_CNSC-RSP-0153-1.pdf.Dose assessment model for chronic atmospheric releases of tritium
[J]. Atomic Energy Sci. Technol., 2010. 44: p. 535 (in Chinese).