2.1 General Description of IMSFR

The standard model of IMSFR is a pure fission fast reactor with a thermal power of 3000 MWth and a mean fuel temperature of 1023 K, which takes reference to the MSFR [3]. The cross and vertical sections of IMSFR are shown in Fig. 1, following with the main parameters in Table 1. There are mainly two partitions in the IMSFR, namely the fuel zone and the fertile zone, corresponding to the fuel salt and the fertile salt. The composition of the fuel salt is set as 22.5(HM)F₄-77.5LiF in mol%. For the Th-U fuel cycle, HM stands for the heavy metal initially composed of ²³²Th and ²³³U, and the ratio of ²³²Th and ²³³U is adjusted for the reactor criticality. The composition of the fertile salt is also 22.5(HM)F₄-77.5LiF in mol%, where the initial HM is only ²³²Th. The proportion of (HM)F₄ is fixed at 22.5 mole% because of the low eutectic point of 838 K. The adopted enrichment of ⁷Li in the MSFR is 99.995%, which will be further studied in this paper.

Fig. 1:Full-screen View

Cross and vertical sections of IMSFR.

The fuel zone (red area shown in Fig. 1) consists of the active core and the external fuel circuit (including collectors, salt-bubble separators, heat exchangers, pumps, and pipes). The active core is a compact cylinder (height/diameter =1) with a diameter of 225.5 cm. The liquid fuel salt flows through the core from the bottom to the top and then to the heat exchangers with 16 groups of pumps outside. At the same time, the heat produced by fission is taken out of the active core to the heat exchangers then to the secondary loop. The total volume of the fuel salt is 18 m³, half of which is in the active core and the other half in the external circuit. The fertile salt surrounds the active core in both the radius direction and the axial direction with a thickness of 30 cm (blue area shown in Fig. 1). The fuel salt and the fertile salt are separated by the Ni-based alloy with a thickness of 2 cm. The activation products of the alloy have been studied in a fluoride salt-cooled high-temperature reactor (FHR) [23], but further study is still needed in the IMSFR to confirm its stability and security. Surrounding the fertile salt is the graphite reflector (yellow area in Fig. 1), which is adopted to the fertile saving [22]. Between the external circuit of the fuel salt and the graphite reflector, B₄C of 10 cm (green area in Fig. 1) is arranged to stop the neutrons leaking out of the reactor. The most external surface is also Ni-based alloy which serves as the structure material. Continuous boundary conditions are adopted between the material of the fuel salt, the alloy, the fertile salt, graphite, and B₄C, while vacuum boundary condition is used outside the most layer out of the alloy.

2.2 Calculation tools: SCALE6.1 and MSR-RS

SCALE6.1 [24] is a comprehensive code package developed and maintained by ORNL, which can be used for criticality safety, burn-up calculation, radiation shielding, radioactive source term analysis, sensitivity and uncertainty analysis, and so on. In this paper, the module of CSAS (Criticality Safety Analysis) is used for the criticality calculation, the TRITON (Transport Rigor Implemented with Time-dependent Operation for Neutronic depletion) module performs for the problem-dependent cross-section processing and multi-group neutron transport calculation, and the ORIGEN-S (Oak Ridge Isotope Generation Code System) module serves for the depletion and decay calculations. The KMART6 (Keno Module for Activity- Reaction Rate Tabulation) module is used for the reaction rate of relevant nuclides. Other modules are also invoked, such as CENTRM/PMC and BONAMI which process the cross sections in the resolved and unresolved resonance energy regions, respectively.

For MSR, the on-line reprocessing is an outstanding feature compared to the traditional solid fuel reactors. The extraction of FPs brings great advantages on the neutron economy and the Th-U fuel breeding. In addition, ²³³U and ²³²Th can be refueled on line for the reactor criticality, which is beneficial to decrease the excess reactivity and the inventory of fissile. However, tools at hand, such as MCNPX and SCALE, are not sufficient to simulate the complex reprocessing, especially for the functions of refueling and extracting with changing rates versus time. So the module of MSR-RS based on SCALE6.1 was developed particularly for the burn-up simulation of MSR in our previous work [25, 26].

The flow chart of MSR-RS is shown in Fig. 2. First, the geometry and the molten salts are initialized for the input files. Second, the neutron transportation and depletion are calculated by TRITON and ORIGEN-S. Meanwhile, FPs are removed and ²³³Pa is extracted as well to decay into ²³³U out of the core. In addition, ²³³U and ²³²Th are injected into the fuel salt in order to maintain the reactor criticality and keep the inventory of HN constant. The cycle calculation is performed iteratively until the depletion time reaches the value set by the user. A 238-group cross-section database is selected in this paper, and 388 nuclides are tracked in the trace quantities.

Figure 2:Full-screen View

Flow chart of MSR-RS.

In every step of the depletion calculations, about 1600 nuclides are tracked including fissions and decays. For the nuclide i, the burn-up equation is shown in Eq. (1):

in which, Ni is the atom density of nuclide i, λi is the decay constant, σ_i is the one group absorption cross section, $\overline{\varphi }$is the average neutron flux, lij is the branch ratio to i when j decays, and fij is the branch ratio to i after j absorbs a nuetron. The last two parameters${\lambda }_i^r{N}_i$and Bi stand for the rates of extracting and refueling, respectively. And${\lambda }_i^r=\in /T_r$is the parameter similar to the decay constant. ϵ stands for the efficiency of reprocessing, and Tr is the period of reprocessing. In this paper, the extraction of FPs and the refueling of ²³³U and ²³²Th are realized by changing these reprocessing parameters in ORIGEN-S.

The on-line reprocessing includes the refueling of HN and the extracting of FPs. The extracting system involves two key processes [3]. One is the mechanical extraction of gaseous FPs and non-dissolved metals via an on-line He gas bubbling system. The other is the chemical removal of lanthanides and other soluble FPs via batch reprocessing by the on-site facilities. The He gas bubbling system can remove the gaseous FPs and the non-dissolved metals rapidly, so the remove period is set as 30 s. The chemical reprocessing is relatively slow and is set as 40 L/d. The fuel salt and the fertile salt are 18 m³ and 8.45 m³, respectively, so their reprocessing periods are 450 d and 210 d. Meanwhile, ²³³Pa and ²³³U in the fertile salt are also extracted along with the molten salt reprocessing, aiming to reduce its neutron absorption and improve the Th-U conversion capability. Because of the on-line refueling which makes it possible for MSR with low excess reactivity, the effective multiplication factor (K_eff) is set at 1.005±0.005 during all the operating time.

3.1 Th-U Breeding Ratio

BR is defined by the product rate over the disappearance rate of fissile, or the capture rate of fertile over the absorption rate of fissile. For Th-U fuel cycle, it can be expressed as [20]:

in which R_a and R_c stand for the neutron capture rate and the absorption rate, respectively.

The results are shown in Fig. 3. From the chart we can see that, when the enrichment of ⁷Li is 99.995%, BR is approximately at 1.11 in equilibrium based on the improved geometry. It is obviously promoted when compared to that in the reference MSFR [3], which results from the contribution of the fertile salt in the axial direction.

Figure 3:Full-screen View

BR’s evolution vs. the enrichment of ⁷Li.

As also can be seen from the chart, when the enrichment is above 99.95%, BR’s evolution with time keeps nearly the same. But when the enrichment is 98%, BR declines drastically to less than 1.0 because of the neutron absorption of ⁶Li. Then, ⁶Li is consumed (more detailed discussion is shown in Sect. 3.4) and its neutron absorption decreases as the evolution of the reactor, leads to an increase of BR. After about 50 years’ evolution, BR, with a different enrichment of ⁷Li, tends to be the same. We also notice that BR declines rapidly at the early time, and it is because of the accumulation of FPs. In short, according to the evolution of BR, the enrichment of ⁷Li could be set at the range between 99.5% and 99.95% rather than 99.995%.

3.2 ²³³U Production and DT

When BR is bigger than 1.0, the fissile produced is more than that consumed. For the IMSFR with on-line reprocessing, the extracted ²³³Pa from the fertile salt decays to ²³³U in half a year. The ²³³U extracted from the core and that decayed from ²³³Pa are partially refueled into the active core for criticality, and the remaining (if any is left) is the net ²³³U production, which is defined as:

where ex and re stand for extracting and refueling, respectively. For more clarity, the relative ²³³U production (M_r) is used in this paper. It is obtained by dividing the net ²³³U production to the initial inventory of ²³³U (M₀), which is expressed as:

That is to say, when M_r reaches 1.0, the net production of ²³³U equals that of M₀ (4.8 t), which means the fissile doubles and it is sufficient to start a new IMSFR. The corresponding time is DT. Hence, the ²³³U production per year can be derived as the following equation:

Fig.4 presents M_r versus time, while Fig. 5 exhibits DT and the ²³³U production per year with the different ⁷Li enrichment. As shown in the two figures, when the enrichment of ⁷Li is 99.995%, the production rate of ²³³U promotes from 95 kg/year in the reference MSFR [3] to 130 kg/year in the IMSFR, while DT shortens from 56 years to 36 years. The advantages of the IMSFR are certified again from the aspects of ²³³U production and DT.

Figure 4:Full-screen View

Relative ²³³U production vs. the enrichment of ⁷Li.

Figure 5:Full-screen View

DT and ²³³U production rate vs. the enrichment of ⁷Li.

As also can be seen from Fig. 4, when the enrichment is above 99.95%, the ²³³U production in different cases is nearly the same. There is no obvious improvement on Th-U fuel breeding when the enrichment is 99.9995% compared to that for 99.95%, and this is the same with DT and the ²³³U production rate as show in Fig. 5. When the enrichment is 99.5%, the ²³³U production declines and the DT extends to more than 40 years. For the case of 98%, the net ²³³U production even could not reach the initial inventory in 80 years. In short, taking into account the ²³³U production and DT, the ⁷Li enrichment should be set at the range between 99.5% and 99.95%. 99.95% is recommended in the IMSFR because it is suitable for both Th-U fuel breeding and a relatively low technical requirement on ⁷Li purification.

3.3 Spectrum

In this section, the normalized lethargy spectral are explored to understand the influences of ⁷Li enrichment on Th-U fuel breeding more deeply.

Figure 6(a) and Fig. 6(b) give the spectra at the initial time with a series of ⁷Li enrichment for the fuel salt and fertile salt, respectively. As can be seen from Fig. 6(a), there is no obvious distinction in the fuel salt with different ⁷Li enrichment. However, when it comes to the fertile salt, significant differences appear, as shown in Fig. 6(b). The lower the ⁷Li enrichment is, the harder the spectrum is. The reason is that the neutrons generated in the active core are moderated during the travelling from the fuel salt to the fertile salt, and the neutron absorption cross section of ⁶Li is much bigger in the low-energy region than in the high-energy region. Therefore, when there is plenty of ⁶Li in the fertile salt, the low-energy neutrons are absorbed, leading to the decline of the spectrum in the low-energy region. The little deviation of spectra in the fuel salt is also because of the absorption of ⁶Li. When the ⁷Li enrichment is higher than 99.95%, there is not enough ⁶Li, so the spectrum keeps nearly the same as the case of 99.9995%. In other words, the ⁷Li enrichment of 99.95% is recommended as well from the aspect of spectrum.

Figure 6:Full-screen View

Normalized lethargy spectrum of fuel salt and fertile salt at initial time with different ⁷Li enrichment (a, b) and spectral evolution with time when the enrichment is 99.5% (c, d).

Then, in the second part of this section, the spectral’s evolution with time is studied in the case that the ⁷Li enrichment is 99.95%. The results are exhibited in Fig. 6(c) and Fig. 6(d), which are also corresponding to the fuel salt and the fertile salt. From the chart we can see that the spectra in both the fuel salt and the fertile salt have almost no change from the startup to the 60th year. Only a little deviation appears which mainly results from the large neutron absorption cross-section of FPs, such as Xe and Sm.

In a word, the enrichment of ⁷Li has a more important influence on spectrum in the fertile salt than that in the fuel salt, which is because the neutron absorption cross section of ⁶Li in the low-energy region is larger than that in the high-energy region.

3.4 Inventory of ⁶Li and Neutron Absorption

Further researches are carried out by analyzing the inventory of ⁶Li and the neutron absorption of the main nuclides, including ⁶Li, ²³²Th, and ²³³U. Fig. 7 gives the evolution of ⁶Li with different enrichments of ⁷Li. As can be seen from the chart, when the enrichment of ⁷Li is higher than 99.95%, the inventory of ⁶Li in the core is less than 5 g, which is so little that it has almost no effect on the thermal neutrons. However, when the enrichment of ⁷Li is 99.5% or less, the mass of ⁶Li is more than 20 g, which significantly affects the neutron absorption as well as the Th-U fuel breeding.

Figure 7:Full-screen View

Inventory of ⁶Li vs. time with different enrichment of ⁷Li.

The relative neutron absorption of the main nuclides versus the enrichment of ⁷Li is given in Fig. 8 at the startup time and at the 36th year (DT is 36 years when the enrichment of ⁷Li is 99.95%). The solid line stands for the neutron absorption at startup. Information can be obtained from the chart that the neutron absorption of ⁶Li declines as the increasing of the ⁷Li enrichment due to the decline of ⁶Li’s inventory. Meanwhile, the absorption of ²³²Th increases while the absorption of ²³³U keeps nearly the same. So according to Eq. (2), BR increases and then tends to equilibrium with the increasing of ⁷Li enrichment. That is to say, when the enrichment of ⁷Li is 99.5% or lower, the neutron absorption of ⁶Li is so high that a certain number of neutrons are wasted, which leads to the reduction of ²³²Th’s neutron capture. When the enrichment is higher than 99.95%, the neutron absorption of ⁶Li is relatively low, and it has little influence on ²³²Th’s neutron capture. The dashed line in Fig. 8 represents the absorption at the 36th year. As shown in the chart, the neutron absorption of ⁶Li, ²³²Th, and ²³³U change a little with the variation of the ⁷Li enrichment. That is because after 36 years’ burn-up, ⁶Li is consumed to a very low level, which has been shown in Fig. 7.

Figure 8:Full-screen View

Relative neutron absorption vs. the enrichment of ⁷Li at startup and at the 36th year

Then, the neutron absorption at startup and at the 36th year are compared in Fig. 8. When the enrichment of ⁷Li is 99.95%, the neutron absorption of ²³²Th and ²³³U is slightly lower at the 36th year than that at startup. That’s because of the accumulation of FPs, which absorbs a certain amount of neutrons in the core. Fortunately, with the on-line reprocessing, the influence of FPs has decreased to a very low level.

3.5 Tritium Production

In the MSRE, it is found that the tritium is mainly from the irradiation of ⁶Li [27], and the generation reaction is as follows:

In the IMSFR with LiF molten salt, the decrease of ⁷Li enrichment would inevitably lead to the increase of ⁶Li inventory and the production of tritium. For the fusion reators, the production of tritium is useful to the D-T fussion reaction, so the increase of ⁶Li in the Flibe or LiF molten salts contributes to the TBR and the tritium self-sufficiency [19]. However, the IMSFR concept is a pure fission fast reactor but not a fusion one. The tritium is extracted as the He gas bubbling because it is undisolved in the molten salt, and following work is needed to collect it. So it is undesired for the possibility of releasing to the atmosphere [28] and should be as little as possible in the fission reactor. In this subsection, the production of tritium is investigated finally.

As shown in Fig. 9, after 80 years’ operation, the tritium production is about 7 kg when the enrichment of ⁷Li is 99.95%, and it is nearly the same to the case of 99.9995%. But when the enrichment of ⁷Li is 99.5%, the production of tritium is more than 11 kg. And it even reaches to nearly 30 kg in the 80th year when the enrichment of ⁷Li is 98%. So when the ⁷Li enrichment of 99.5% is still recommended from the aspect of tritium production.

Figure 9:Full-screen View

Tritium production with different enrichment of ⁷Li.