1. Introduction

The concept of the molten salt reactor originated from Oak Ridge National Laboratory in the 1950s [1] using fluoride salts as both a coolant and fuel. In 2003, a fluoride salt-cooled high-temperature reactor concept, or FHR, was proposed by American scientists [2, 3]. The Center for Thorium Molten Salt Reactor System (TMSR) of the Chinese Academy Sciences (CAS) adopted the FHR concept and developed conceptual designs for test reactors in 2012 [4, 5]. As with other FHRs, the TMSR’s design uses the TRISO (Tri-structural Isotropic) coated-particle fuel, molten FLiBe (2LiF-BeF₂) salt coolant, a graphite reflector, and alloy-N structural material, which has a high inherent safety under the design basis accidents, such as LOOP (Loss Of Offsite Power) accidents [6] and even SBO-ATWS (Station Blackout Anticipated Transient Without Scram) accidents [7].

The activation of the alloy-N by fission neutrons has a significant impact on the maintenance and decommissioning of the reactor. The alloy-N may be corroded and dissolved in the molten FLiBe salt and then irradiated by fission neutrons when the salt flows through the core. The alloy-N inside the reactor vessel will be activated by fission neutrons directly and then corroded into the molten FLiBe salt. The activated alloy-N in the salt may be deposited back onto the inner surface of the primary loop components, which is made of alloy-N. The primary loop components, containing long-life activation products that emit high-energy γ-rays, pose threats to the workers’ health during maintenance and decommissioning. The activation of the primary loop salt has been studied before [8, 9].

The corrosion and deposition of alloy-N in the FLiBe salt is a complicated process involving many physical and chemical mechanisms such as intrinsic corrosion, corrosion by oxidizing contaminants, differential solubility, galvanic corrosion, and so on [10].These topics are well beyond the scope of this paper. Instead, this paper focuses on analyzing the activation of unit mass alloy-N in the core. Once the corrosion process is characterized correctly by future works, the results of this paper can be used to calculate the radiation source terms of the components and salt to support radiation protection, operation, maintenance, and decommissioning planning.

2. The Reactor Core

The core of the TMSR’s test reactor is an octagonal prism filled with fuel pebbles and graphite pebbles. The graphite pebbles are only at the bottom and the top of the core. Each fuel pebble has a diameter of 6cm and contains an average of 11,660 TRISO particles in its fuel zone. Each fuel pebble contains 7g of uranium and the ²³⁵U enrichment is 17 wt %. Each TRISO particle consists of a UO₂ fuel kernel and several coating layers. The TRISO particles are randomly dispersed in a graphite matrix fuel zone in the pebble. A hard graphite shell encloses the fuel zone.

The pebble bed is surrounded by graphite reflectors with an outer diameter of 260cm and a height of 300cm. The thicknesses of the graphite reflectors on the top of the core and below the core are 65 cm and 50cm, respectively. Channels are distributed in the top and bottom reflectors to allow the molten salt to flow through. Channels for situating control rods, measurement instruments, and neutron sources are located in the radial reflectors. The design parameters are listed in Table 1 [4].

Alloy-N is a 70.01%Ni-16.50%Mo-7.03%Cr nickel-based alloy whose nominal composition is shown in Table 2 [11]. Ni and Mo in alloy-N provide excellent resistance to molten fluoride salt corrosion. However, the other metallic alloying elements in alloy-N may suffer from corrosion [12, 13]. Since this paper focuses on analyzing the neutron activation in the core, a unit mass of one gram of alloy-N is used in the calculation. Alloy-N is assumed to be in the coolant and irradiated by the fission neutrons. After the irradiation, the activities of alloy-N are primarily due to the short-lived ²⁸Al, ^60mCo, ⁵⁶Mn, ⁵¹Ti, and ⁵²V, as well as the long-lived ⁶⁰Co, ⁵¹Cr, ⁵⁵Fe, ⁵⁹Fe, and ⁵⁴Mn. Table 3 shows the properties of these nuclides[14, 15].

3 SCALE Model of the Core

In the multi-group (MG) neutron transport calculations, the initial problem-independent MG libraries were generated from the Evaluated Nuclear Data File (ENDF/B) using a generic flux spectrum. Before being used in the MG neutron transport calculation, the problem-independent MG library must be corrected for space-dependent and resonance self-shielding effects based on the unit cell description. FHR, which uses the fuel pebble, has double heterogeneity. The TRISO particles embedded in the graphite matrix constitute the first level of heterogeneity and the fuel pebbles with the moderator and reflector form the second level heterogeneity. This double heterogeneity must be treated through the unit cell description in order to obtain the accurate problem-dependent MG library.

We have built a SCALE 6.1 [16] computer model for the reactor core in order to calculate the activation products of alloy-N. The main challenge in modeling the core is to properly treat the double heterogeneity, which can be completed by the DOUBLEHET module. According to the unit cell described by the DOUBLEHET, SCALE uses the conventional Bondarenko method to process the shielding factors in the unresolved resonance range and a deterministic point-wise calculation of the fine-structure spectra in the resolved resonance and thermal energy ranges to generate the problem-dependent MG cross sections, which are then used to perform the depletion analysis.

The repeated structural unit chosen to build the active core consists of a pebble at the body-center of the cubic and a 1/8 pebble at every corner of the cubic. Figure 1 shows the repeated structural unit, which has the same length and width of 7 cm and a height of 6.78 cm, which maintains a pebble packing factor of 64%. These repeated structural units are arranged in an infinite square lattice and enclosed by an octagonal prism to form the active core. The pebbles in contact with the surfaces of the octagonal prism may be cut. Several types of special structural units have been used to replace the repeated structural units, which stay within the boundary of the core to ensure one whole pebble stays within the boundary.

Figure 1Full-screen View

(Color online) The repeated structural unit to form the active core

A model of the core is shown in Figure 2. The shutdown and control system channels are included in the model but filled with a vacuum. That is to say, the control and shutdown rods are assumed to be fully withdrawn from the core. The boundary condition of the model is set to be a vacuum.

Figure 2Full-screen View

(Color online)Model of the core: (a) cross section; (b) three-quarters of the total profile

The control module, TRITON, of SCALE 6.1 is used to calculate the activity of the activation products. KENO-VI performs neutron transport analysis, and is coupled with the ORIGEN module to perform the depletion calculations. The cross section library of v7-238 is used and the CENTRM module is used for cross section processing.

4. Results and Discussions

4.1 Neutron spectrum

The activation products, such as ²⁸Al, ^60mCo, ⁵⁶Mn, ⁵¹Ti, ⁶⁰Co, ⁵¹Cr, ⁵⁵Fe, and ⁵⁹Fe, are produced by thermal neutron capture reactions. However, ⁵⁴Mn and ⁵²V are produced by fast neutron (n, p) reactions. Thus, the neutron spectrum may significantly affect the production of the activation products. Since the alloy-N is dissolved in the coolant, the neutron spectrum averaged over the coolant in the core is of great interest, and the neutron spectrum at the beginning of the cycle is shown in Figure 3. The spectrum is calculated using the model described in section 3. The reactor has a thermal neutron spectrum and the ratio of thermal flux to total flux is approximately 40%.

Figure 3Full-screen View

Neutron spectrum averaged over the coolant in the core at the beginning of the cycle.

4.2 Activity of alloy-N

4.2.1 Activity in power operation

Activities of the alloy-N per unit mass after the reactor runs at full power for 180 days are listed in Table 4. Activities of each activation product changing with the irradiation time are shown in Figure 4. Short-lived nuclides such as²⁸Al, ^60mCo, ⁵⁶Mn, and ⁵¹Ti reach their maximum activities before the end of the cycle due to their very short half-life. However, ⁵²V, a very short-lived nuclide, does not reach its maximum activity immediately because of the small activation cross-section of ⁵²Cr to produce ⁵²V. Long-lived nuclides such as ⁶⁰Co, ⁵¹Cr, ⁵⁵Fe, ⁵⁹Fe, and ⁵⁴Mn reach their maximum activities at the end of the cycle. The nuclides of ⁵⁶Mn, ⁶⁰Co, ²⁸Al, ⁵⁹Fe, ⁵²V, and ⁵⁴Mn undergo β^- and γ decay by emitting high-energy γ-rays. Thus, in the power operation of the reactor, ⁵⁶Mn, ⁶⁰Co, ²⁸Al, ⁵⁹Fe, ⁵²V, and ⁵⁴Mn may be important in terms of radiation protection.

Table 4Full-screen View

Activities of alloy-N per unit massat the end of the cycle

Nuclide	ActivityBq/g	Nuclide	ActivityBq/g
56Mn	1.33E+10	28Al	1.70E+08
51Cr	9.20E+10	59Fe	2.70E+07
60mCo	9.15E+09	52V	1.18E+07
60Co	1.00E+09	54Mn	2.78E+06
55Fe	1.16E+08	51Ti	3.16E+06

Show more

Figure 4Full-screen View

(Color online)Activities of the activation products changing with the irradiation time

4.2.2 Activity after shutdown

After the shutdown of the reactor, the activity decreases due to decay. Activities decreasing with the cooling time are shown in Figure 5. The short-lived nuclides of ²⁸Al, ^60mCo, ⁵⁶Mn, ⁵¹Ti, and ⁵²V die out in a few days, and the long-term radiation source terms are primarily due to the long-lived nuclides of ⁶⁰Co, ⁵⁵Fe, ⁵⁴Mn, ⁵¹Cr, and ⁵⁹Fe. However, ⁵⁵Fe and ⁵¹Cr undergo electron capture and emit very low-energy x-rays and γ-rays. Thus, the radiation source terms of the irradiated alloy-N are mainly due to ⁶⁰Co, ⁵⁴Mn, and ⁵⁹Fe. The nuclides of ⁶⁰Co, with a half-life of 5.72 years, undergo β^-decay and emit two γ-rays with energies of 1.17MeV and 1.33 MeV, respectively. The nuclides of ⁵⁴Mn, with a half-life of 312 days, undergo electron capture and emit 0.835 MeV γ-rays. The nuclides of ⁵⁹Fe, with a half-life of 44.63 days, undergo β^-decay and emit two γ-rays with energies of 1.099 MeV and 1.292 MeV, respectively. Furthermore, due to their long half-life, the activities of ⁶⁰Co and ⁵⁴Mn increase with the irradiation time almost linearly. The production rates of ⁶⁰Co and ⁵⁴Mn per unit mass of alloy-N are 5.58×10⁵ Bq∙MWd^-1∙g^-1 and 1.55×10³Bq∙MWd^-1∙g^-1, respectively.

Figure 5Full-screen View

(Color online)Activities of the activation products after shutdown

In addition, attention should be paid to ^60mCo. ^60mCo undergoes isomeric transition and β^-decay with a branching ratio of 99.76% and 0.24%, respectively. Its isomeric transition produces ⁶⁰Co, which is a main contributor to the radiation source terms.

5. Conclusion

We have constructed a SCALE 6.1 model for the TMSR’s test reactor core and used the model to calculate the average neutron spectrum in the coolant. We also used the model to calculate and analyze the activity of each constituent in the irradiated alloy-N. The results show that among the activation products, ⁶⁰Co and ⁵⁴Mn, which emit high-energy γ-rays, are the major contributors to the radiation source terms considering their relatively long half-life. Because of the corrosion of alloy-N by the FLiBe salt, the activation products of the alloy-N may be dissolved in the coolant and deposited onto different components of the primary loop. The exact corrosion and mass transport process is yet to be understood. Nevertheless, the salt and components may be highly radioactive after a prolong power operation, even after a long cooling time, because of the existence of ⁶⁰Co. This is a unique situation for FHRs and may complicate maintenance and decommissioning. Therefore, it is important to quantify the activation of alloy-N in the primary loop. This paper is the first step in analyzing the activation of alloy-N in FHRs.