19. Transmutation Scenarios after Closing Nuclear Power Plants

An ADS core dedicated for MA transmutation (MA-ADS) [8] bases the present design for Pu transmutation. As listed in Table 19.1, the MA-ADS is a medium-size core loaded with nitride fuel cooled by lead-bismuth eutectic (LBE). Nitride fuel is diluted by zirconium-nitride with a weight fraction of around 50 % (volume fraction around 65 %) that can be changed so that criticality becomes an appropriate level (k _eff = 0.97). The ADS is operated for 600 effective full-power days (EFPDs) without any fuel reloading while the interval for maintenance of an accelerator can occur. In other words, the ADS is a one-batch core that implies large reactivity drop after depletion in the case of Pu transmutation. Because control rods for criticality are not equipped in the ADS, the drop must be supplemented by increasing proton beam current. Fortunately, MAs are fertile material that becomes fissile after capturing one neutron, and the drop is very small, as shown in the following section. Table 19.3 provides weight composition of MA fed to MA-ADS. Figure 19.2 illustrates the R-Z model for calculation.

Two cases of Pu feed were assumed in the present design for Pu transmutation: Pu-ADS, to which only Pu is provided from reprocessing process of LWR SF, and Pu+U-ADS, to which Pu accompanied by U with 50 % weight ratio is provided from the process. Treatment of pure Pu raises proliferation concern in a country without nuclear weapons, and the reprocessing plant is designed to add depleted U to Pu just after separation. Addition of ²³⁸U results in ²³⁹Pu production and generally is undesirable for Pu transmutation. Table 19.3 lists two compositions. Other conditions are similar to the MA-ADS.

The reference ADS for MA transmutation is designed with a one-batch core, which means that all the fuel is loaded and unloaded simultaneously. At first, this one-batch design was adopted to Pu transmuters. Table 19.4 lists the in-core and out-core time with operation efficiency. Out-core time of the ADSs is 3 years, which allows decay of ²⁴⁴Cm. In-core time of the reference MA-ADS is 2 years, although that of Pu-ADS is reduced to 1 year because the decrease of criticality is too rapid for this ADS. The operation efficiency, ε _o, is 82.1 % assuming 300 days operation annually.

Design and transmutation performance are summarized in Table 19.5. Volume fraction of the inert matrix, ZrN, of the MA-ADS core is 69.8 %, adjusted so that k-effective at the beginning of the cycle (BOC) of the equilibrium core becomes 0.97. The equilibrium core is obtained after calculating ten cycles of burning, cooling, and recycling. Volume fraction of the Pu-ADS is more and that of the Pu+U-ADS is almost the same. The inventory at BOC of the heavy metal in Table 19.5 is proportional to a one-volume fraction of ZrN. An interesting observation is that the amounts of Pu at BOC are equal among three ADSs, which means U and MA contribute very little to the criticality before depletion. However, impacts on the criticality drop after depletion is significant (Fig. 19.3). k_eff drop of the MA-ADS is as small as 1.5 %dk, although others lose 14 %dk even at the equilibrium cycle around 6,000 days, which means MA is a better fertile than ²³⁸U. The Pu-ADS has a steeper decrease than Pu+U-ADS because of the absence of ²³⁸U. The huge drop of the Pu- and Pu+U-ADS is not acceptable in the current design of accelerator and target for the MA-ADS; the acceptable drop is about 3 %dk in the MA-ADS.

The effective transmutation rate and transmutation half-life are listed at the bottom of Table 19.5. The half-life of the Pu-ADS is shortest because its specific heat is twofold larger than others although its cycle efficiency, ε _c, is much smaller than others.

In the one-batch design in the previous section, k_eff drop of Pu ADSs is 14 %dk, which is too large to be compensated by burnable poison or control rods. As the first step of design improvement, a multi-batch design is introduced. Theoretically, an N-batch core can reduce k_eff drop by 1/N, that is, a drop of 14 %dk can be reduced to 2.3 %dk by six-batch design. Figure 19.4 illustrates the criticality change with an expansion for operation date of 0–1,200 days. The criticality drop for the early operation date is larger than the limit of 3 %dk, that is, 0.94 of k_eff at end of burn-up; the drop decreases in the equilibrium cycle. The maximum drop is 5 %dk for Pu-ADS and 6.5 %dk for Pu+U ADS, which can be compensated by control rods or burnable poison or mitigated by shorter operation in the early cycle in future improvements. The drop in the equilibrium cycle is approximately 2 %dk, which is comparable to that of the reference MA-ADS.

Volume fractions and inventories are listed in Table 19.6. Six-batch cores generally require more inventory than a one-batch core because an averaged k_eff during operation of multi-batch cores is higher than that of the one-batch core. Transmutation amounts of Pu- and Pu+U-ADSs are much smaller than that of the MA-ADS because the operation period is, respectively, only 50 and 100 days.

To evaluate transmutation half-life, operation and cycle efficiencies must be determined. The short operation period of 50 or 100 days implies frequent fuel exchange and low operation efficiency. There are two kinds of interval: fuel exchange and plant maintenance. We assumed that fuel exchange of a 1/6 core requires 15, 30, or 60 days for the Pu- and Pu+U-ADS and that plant maintenance including accelerator needs 60 days. Because fuel exchange for 15 days is very short, considering shutdown and startup of the ADS plant is included, tentative storage inside a core vessel should be applied for such a short interval. In the case of Pu-ADS, the 50-day operation and 15-, 30-, or 60-day interval are repeated five times, then 50-day operation and 60-day maintenance are done. In the case of Pu+U-ADS, 100-day operation and 15-, 30-, or 60-day interval are repeated two times, then 100-day operation and 60-day maintenance are done. The total operation period before a long plant maintenance of 60 days in both ADSs is 300 days.

Based on the foregoing assumptions, operation efficiency and cycle efficiency are determined as listed in Table 19.7, with specific heat and resulting transmutation half-life. Operation efficiency multiplied by cycle efficiency of the Pu-ADS is the poorest, but the transmutation half-life is the shortest because of the high specific heat. In the present study, a 30-day interval for fuel exchange is adopted as a nominal case. The transmutation half-life of the Pu-ADS is 24.8 years in the nominal case, which is applied to scenario analysis.

Another observation is that the impact of the out-core period on cycle efficiency is significant. The out-core period is presumed considering the half-life of ²⁴²Cm of 126.8 days. If a shorter out-core period is accomplished by corresponding design of the reprocessing and fabrication plant, cycle efficiency and resulted transmutation half-life can be improved. Table 19.8 shows comparison of a 3-year and 1-year out-core period. An impact on the transmutation half-life of the Pu-ADS is a factor of around 2, and the transmutation half-life becomes as short as 13.5 years. Although 3 years of out-core period is applied as the nominal case, a shorter out-core period should be pursued in future study.

Figure 19.5 illustrates a result of the LWR-OT scenario where time evolution of electricity generation, Pu inventory, and MA inventory are shown. The peak of 50 GWe appears in 2010 and decreases because of the Fukushima accident and closure after 40-year operations. All LWRs will be shut down in 2055. The Rokkasho reprocessing plant (RRP) will not be operated, but 7,100 tHM spent fuel has been reprocessed, mainly overseas. The year of reprocessing is not clear but assumed to be in the 1990s. A small amount of MOX fuel from this reprocessing will be utilized in LWRs.

Pu inventory mainly exists in UO₂-SF. “Pu” in the figure is not “separated” Pu, but Pu in MOX fresh fuel in this scenario. The total of plutonium is 350 t that is gradually disposed of to a repository from 2043 until 2105. The trend of MA inventory is almost the same, but it continues to increase after 2040 because ²⁴¹Pu becomes ²⁴¹Am with a half-life of 14.35 years.

In MOX scenarios, the Rokkasho reprocessing plant (RRP) will be operated with annual capacity of 800 t and a MOX fabrication plant also (Fig. 19.6). The total amount of UO₂-SF reprocessed is 34,500 tHM, slightly larger than the planned amount of 32,000 tHM. Thus, the present analysis assumes an extension of the RRP by several years. The MOX loading to a usual LWR is limited to 30 %, although the Ohma full-MOX reactor starting in 2014 in this analysis can be operated only by MOX fuel. Because the RRP is operated after all LWRs are closed, part of the separated Pu cannot be burned. The total amount of Pu is reduced to 250 t, but that of MA is increased to 100 t.

In the FR scenario as well as other transmutation scenarios, Pu from the RRP is at first fabricated as LWR-MOX fuel and burned in LWR. Pu is co-extracted with same content of U in the current RRP, although MA is vitrified as waste. MA partitioning is assumed to be introduced in 2025 and stored until 2045. In 2045, before introduction of transmuters in 2050, reprocessing of LWR-MOX spent fuel will begin and provide Pu to the transmuters.

FRs are to be introduced in 2050 when 250 t plutonium and 100 t MA remains. MA of 20 t is vitrified by the RRP before 2025 and is not available for transmutation. Available TRU is 330 t. The required TRU to introduce an FR is approximately 25 t, if we assume 41 % of Pu content and 15 % of MA content and employ 45.1 t from Table 19.10. Theoretically, 14 (=350/25) FRs can be introduced in 2050, but only 8 can be deployed in practice because the plant life of an FR is assumed to be 60 years and sufficient TRU must be kept until 2110. Available TRU gradually decreases to 200 t in 2110 by transmutation. After 2110, FRs are replaced and reduced to 3 units corresponding to available TRU of 200 t that decreases to 130 t in 2170. Then, 2 FRs from 2170 to 2230 and 2 FRs from 2230 to 2290 will be deployed. After four generations of transmutation, the amounts of Pu and MA are reduced to 40 and 30 t, respectively.

MA content of FR is as high as 15 % (Fig. 19.7), which is above the design limit of 5 % in Wakabayashi et al. [1]. In the usual design of FBRs, MA accumulation is mitigated by a supply of fresh Pu from the blanket. Moreover, high Pu content of FR burner contributes to high MA content. High MA content generally causes deterioration of safety parameters (beta, Doppler coefficient, void reactivity) and difficulty in a reprocessing and fabrication plant.

In the ADS scenario, transmuter is changed from FR to ADS. ADS can accept both Pu and MA; distribution is shown in Fig. 19.8. In 2050, 22 ADSs are to be introduced, corresponding to 140 t available TRU. Then, 7 and 3 ADSs are operated respectively from 2110 to 2170 and 2170 to 2230. After three generations, Pu and MA are reduced to 10 t and 3 t, respectively, excluding 16 t MA in vitrified waste.

In the FR+ADS scenario, MA content in FR is limited below 5 % with respect to design limit and the remaining MA is transmuted in the ADS. In the first generation of transmutation from 2050 to 2110, six FRs and three ADSs are deployed, then three FRs and two ADSs in the second generation, and two FRs and one ADS in the third generation are built (Fig. 19.9). In the fourth generation, only ADS is utilized as to reduce TRU rapidly. The total amount of Pu and MA is reduced to 20 and 10 t, respectively, excepting MA in vitrified waste.

One of the impacts on the repository by transmutation is reduction of potential radiotoxicity, which is defined as total ingestion dose of the waste. Because waste is isolated from the public in the underground in reality, such direct ingestion never occurs and it is considered to be hypothetical, but it can represent the potential danger of waste. This toxicity of waste can be compared to that of uranium ore consumed for electricity generation causing radioactive wastes. Figure 19.10 illustrates those toxicities corresponding to whole operation of LWRs and transmuters. Consumed natural uranium is 370,000 t.

When wastes are generated, the toxicity becomes higher than corresponding uranium ore by three orders of magnitude. Fission products such as Sr and Cs are dominant in the early several hundreds of years, although actinides contribute to toxicity after that. Toxicity in the LWR-OT scenario decays to the level of uranium ore after 100,000 years. By reducing Pu in the LWR-PuT scenario, the decay time becomes shorter, to 70,000 years. In the transmutation scenarios, shortening of decay time depends on the remaining amount of TRU. The decay time is about 10,000 years in the ADS scenario in which the remaining TRU is approximately 30 t, including vitrified wastes. In comparison between the LWR-OT scenario and the ADS scenario, the amount of TRU is reduced by one order of magnitude, so toxicity is also reduced by same order. If MA in the vitrified wastes is retrievable, the amount of TRU will be reduced to around 10 t, which implies toxicity is reduced to 1/30 and the decay time is around 2,000 years. Thus, the impact on toxicity by transmutation is significantly affected by MA in the vitrified wastes. Early introduction of MA partitioning to the RRP and R&D for retrievability from the glass wastes is of importance in this aspect.

Another impact on the repository is reduction of repository size by partitioning and transmutation of heat-generating nuclides in the wastes. Repository size is represented by a repository footprint, which is defined as an area devoted for waste excluding aisles, ducts, utility area, surface facility, and other.

In the LWR-OT scenario, the footprint corresponding to 45,000 t spent fuel reaches almost 4 km², which is double the typical repository design for the glass waste corresponding to 40-year operation of the RRP, because the spent fuel assembly occupies more area and heat generation from the Pu in it also contributes.

In the LWR-PuT scenario, two kinds of waste form are produced: 37,000 glass waste forms containing FP and MA, and spent fuel assembly of MOX of 4,000 t. Each occupies 1.6 km², and the total is 3.3 km². Although an amount of MOX spent fuel is smaller than that of UO₂ spent fuel in the LWR-OT scenario by a factor of 11, it contains more heat-generating actinides such as Am and Pu, and its footprint is significant.

In the early several hundreds of years, ⁹⁰Sr and ¹³⁷Cs, whose half-life is around 30 years, are dominant for the footprint. They are separated in the RRP after 2025 as well as MA in the transmutation scenarios. They are absorbed by adsorbents such as zeolite and calcined to the waste form. Because half-life is rather short and the repository footprint is almost proportional to heat generation, long-term storage of the calcined waste is effective [9]. After 300 years of storage, an accumulated layout for the TRU wastes that is low heat generating and with long-term radioactive wastes becomes available. The footprint of this layout is smaller by two orders of magnitude than a typical layout for the vitrified waste. After separating ⁹⁰Sr and ¹³⁷Cs, ²⁴¹Am, whose half-life is 432.2 years, becomes dominant, but this nuclide is transmuted in the transmutation scenarios. Heat generation from other fission products that are vitrified quickly decays to the level of the TRU waste.

As result of the long-term storage and transmutation, the footprint becomes almost constant after 2025 (Fig. 19.11). The glass waste form that is produced before 2025 and contains MA occupies 0.5 km². In the ADS scenario, partitioning and long-term storage of Sr and Cs in the wastes produced from reprocessing of ADS spent fuel is not assumed because the impact is small. As a result, the footprint gradually increases to 0.8 km². Technologically, separation is possible in the reprocessing for ADS, and it will be applied if the increase becomes significant. Steps observed in 2230 and 2330 are caused by wastes of remaining TRU that will be diluted to the glass waste, considering heat generation. The remaining TRU of the FR scenarios are more than that of the ADS scenario.

In the transmutation scenarios, the final footprint is around 0.8 km², which is a fifth of the LWR-OT scenario. As is the case of radiotoxicity, the time of introducing partitioning is significant because more than half of the repository is occupied by glass waste forms with MA.

Table 19.11 summarizes the results of scenario analysis. In comparison between LWR-OT and LWR-PuT, reductions are observed in Pu amount, repository footprint, and decay time of toxicity, although they are not drastic. Important benefits of MOX utilization are Pu isotopic deterioration as a nuclear weapon and improved confinement of radionuclides by calcinations, as discussed by Nishihara et al. [7]. However, there remains 110 t of separated Pu that can raise concerns about proliferation.

In comparison between conventional and transmutation scenarios, significant reductions of TRU amount, repository area, and decay time of toxicity are observed. The remaining Pu undergoes several irradiations in the transmuter and is highly resistant to weapon utilization. Repository area is about one fifth and decay time is reduced by one tenth in the maximum case. To achieve such benefit, a total of 15–32 transmuters have to be introduced for 180–240 years with corresponding reprocessing and fabrication facilities. Cost and risk during operation of these facilities would be high compared to their reduction in the repository in the further future.

Comparing the transmutation scenarios, the number of units in the FR scenario is fewest owing to its high thermal output. However, transmutation performance is less than the ADS, resulting in longer transmutation era and larger remaining TRU. Figure 19.12 shows amounts of TRU excluding vitrified MA. An amount of the ADS scenario is reduced to 125 t after the first generation of the transmutation era from 2050 to 2110. In the FBR scenario, equal decrement is achieved after the second generation. Very high MA content up to 15 % in the FR fuel is also problematic. The decrement of the FR+ADS scenario is as same as that of the FR scenario until the second generation, but after that the decrement becomes faster because of ADS introduction.

Although transmutation performance of the FR is inferior to the ADS, cost including profit of electricity generation by transmuter would be much smaller than the ADS because the number of ADSs is doubled, accelerator cost is added, and thermal efficiency of the ADS is worse. Considering high MA content in the FR scenario, the FR+ADS scenario can be a modest solution, although the ADS scenario is preferable if rapid transmutation is required regardless of the cost.