Annals of Nuclear Energy 111 (2018) 635–643

Contents lists available at ScienceDirect

Annals of Nuclear Energy

journal homepage: www.elsevier.com/locate/anucene

Fuel cycles optimization of nuclear power industry in China

Qiang Yue a,⇑, Jingke He a, Shengke Zhi b, Hui Dong a
a
State Environmental Protection Key Laboratory of Eco-Industry, Northeastern University, Shenyang 110819, China
b
Clean Energy, AMEC, Knutsford WA16 8QZ, UK

a r t i c l e i n f o a b s t r a c t

Article history: With the rapid rise of installed nuclear power in China, meeting the increasing demands on natural ura-
Received 11 July 2017 nium and rationally treating the vast spent fuel are essential issues for the sustainable development of
Received in revised form 24 September Chinese nuclear power industry. This paper discusses four most potential nuclear fuel cycle modes in
2017
China and analyzes the natural uranium requirements under these different fuel cycle modes first based
Accepted 28 September 2017
Available online 7 October 2017
on three development patterns (low-, medium-, and high-speed) of installed nuclear power capacity.
Then, an optimization model including natural uranium requirements, spent fuel final disposal amounts
and total cost of electricity generation is constructed and optimization problem under two scenarios of
Keywords:
Nuclear power industry
reprocessing capacity are solved and results discussed. The annual and cumulative natural uranium
Fuel cycle requirements under these two scenarios are also calculated. Finally some conclusions are put forward
Optimization based on the analyses.
Natural uranium Ó 2017 Elsevier Ltd. All rights reserved.
Spent fuel
Cost

1. Introduction resources use efficiency is closely related to the modes of nuclear

Continually growth of energy demand with the development of
economy, the shortage of fossil fuels supply and the limits on
greenhouse gas emissions will greatly speed up the nuclear power
industry’s development in China in the coming next decades
(National Development and Reform Commission, 2007; China
Academy of Engineering, 2011). At the same time as effected by
the Fukushima accident, China’s nuclear power development is
slower than expected in recent years. According to some plans
and research reports, it is reasonable to present three scenarios
for nuclear power development in China to 2050 (see Table 1).
It can be seen from Table 1 that nuclear power will step into a
rapid development period in the future in China. To promote the
sustainable development of nuclear power, the sustainable supply
and use of nuclear fuel, as well as the treatment of spent fuel, are
very important issues for the nuclear power industry. However, the
actual state of China’s uranium resources is not ideal. Uranium ores
have been predominantly low-grade, with 0.05%–0.3% grade ore
accounting for the highest percentage. At the same time, the
reserves were mainly small- and medium-sized (accounting for
more than 60% of the total reserves) (China Academy of
Engineering, 2011). Enhancing the uranium use efficiency is an
important issue for the nuclear power industry in China. Uranium
⇑ Corresponding author.
E-mail address: yueq@smm.neu.edu.cn (Q. Yue).
fuel cycle, so the development of nuclear fuel cycle modes have a
vital role for the sustainable development of nuclear power
industry.
Some studies on the development of nuclear fuel cycle modes
have been carried out. Kunsch and Teghem (1987) carried out
the optimization analysis of nuclear fuel cycle modes by using
multi-objective stochastic linear programming application, con-
forming to the four standards of production cost, resource supply,
balance of business and provide employment as the goals. Kim
et al. (1999) compared the six kinds of fuel cycle [pressurised
water reactor(recycling uranium, Pu), pressurised water reactor
(mixed oxide)-pressurised heavy water reactor(recycling ura-
nium), pressurised heavy water reactor(Pu-U), pressurised heavy
water reactor(direct use of spent PWR fuel in CANDU reactors),
pressurised water reactor and pressurised heavy water reactor
(once through)] by Goal Programming Method and Analytic Hierar-
chy Process (AHP), considering the four quantifiable factors: the
fuel demand, the total depreciation cost, cost sensitivity and the
environmental impact, the conclusion is the adoption of pres-
surised water reactor(mixed oxide)-pressurised heavy water reac-
tor(recycling uranium) is the optimization scheme. Liu et al. (2006)
forecasted the natural uranium resources demand and spent fuel
generation for PWR in China till 2035, and discussed the effects
of spent fuel reprocessing and plutonium separating to reduce
the natural uranium demand and waste accumulation of China’s
nuclear power industry in the future. Bernard (2007) proved mixed

https://doi.org/10.1016/j.anucene.2017.09.049
0306-4549/Ó 2017 Elsevier Ltd. All rights reserved.
636 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643
Table 1
Installed nuclear power capacity in China to 2050 (unit: GWe).
Year 2020
Capacity Low Medium High
40 58 70
Sources: 1. China Academy of Engineering (2011); 2. National Development and Reform Commission (2007); 3. OECD Nuclear Energy Agency and International Atomic Energy
Agency (2014); 4. Nuclear Energy Agency and International Energy Agency (2015).
oxide (MOX) fuel and uranium oxide (UOX) fuel are very similar by
the French thermal reactor operating experience, the use of MOX
can reduce almost equivalent amount of UOX. Cao et al. (2013)
given China’s nuclear power installed capacity in 2050 in accor-
dance with the three different growth speed (300 GW, 400 GW,
500 GW) according to the present situation of China’s nuclear
power development and long-term development plan, and fore-
casted natural uranium resources demand, the amount of spent
fuel, plutonium and minor actinides (MA) under these three
schemes by 2050, and pointed out that advanced nuclear fuel cycle
mode and spent fuel reprocessing plant construction can effec-
tively reduce the accumulation of spent fuel quantity. Worrall
(2013) analyzed the utilization of used nuclear fuel in a potential
future US fuel cycle scenario. Ma et al. (2013) calculated the natu-
ral uranium demand and spent fuel accumulated quantity under
‘‘once through” fuel cycle mode and closed fuel cycle mode by
using the dynamic analysis program of DESAE-2, the results
showed that the closed fuel cycle mode has certain advantages in
the saving of uranium resources and reducing the amount of spent
fuel disposition compared with the ‘‘once through” fuel cycle
mode. Yolanda Moratilla Soria et al. (2015) analyzed the impact
of the taxes on used nuclear fuel on the fuel cycle economics in
Spain. Zhang et al. (2016) carried out the uranium demand and
economic analysis of different nuclear fuel cycles (the once-
through cycle route and the partial recycling in PWR route) in
China. Park (2017) assessed the spent nuclear fuel amounts to be
managed based on disposal option in Republic of Korea. Yoon
et al. (2017) used an integrated multicriteria decision-making
approach for evaluating nuclear fuel cycle systems for long-term
sustainability on the basis of an equilibrium model.
However, few studies have been carried out on fuel cycles
modes optimization in China’s nuclear power industry. This paper
will give four most potential nuclear fuel cycle modes and analyze
the natural uranium requirements under these different fuel cycle
modes first. Then an optimization model including natural ura-
nium requirements, final spent fuel disposal amounts and total
cost of electricity generation will be constructed and optimization
problem will be solved and results discussed. Finally some conclu-
sions will be put forwarded based on the analyses.
2. Natural uranium requirements under different nuclear fuel
cycle modes in China
2.1. Nuclear fuel cycle modes
Pressurised water reactor (PWR) and pressurized heavy water
reactor (PHWR) are the two types of reactors used in China now.
Sodium fast reactors are expected to be put into use in the 2030s
in China. So four most potential nuclear fuel cycle modes of China’s
nuclear power industry in the future are mainly discussed here
(see Fig. 1).
(1) PWR and no recycling of spent fuel [PWR-OT (Once
Through)]
(2) PHWR and recycling uranium (RU) used in PHWR (PHWR-
RU)
2030 2050
Low Medium High Low Medium High
83.8 150 200 250 325 400
(3) PWR and mixed oxide (MOX) used in PWR (PWR-MOX)
(4) Sodium fast reactor (SFR) and MOX used in SFR (SFR-MOX)
2.2. Scenario analysis
Natural uranium requirements under different nuclear fuel
cycle modes till the year 2050 will be analyzed.
2.2.1. Scenario I: The newly-built reactors are all PWRs and no
recycling of spent fuel (PWR-OT)
The reactors in operation now in China are two PHWRs and the
installed capacity is 1456 MW, the other reactors are all PWRs. If
the newly-built reactors are all PWRs in China till the year 2050,
then natural uranium requirements under this condition will be
as the following
MNU ¼ a1  IC PWR þ a2  IC PHWR ð1Þ
where MNU is annual natural uranium requirements, t/a; a1 and a2
are natural uranium requirements per power rating of PWR and
PHWR, t/MW; ICPWR and ICPHWR are installed capacity of PWR and
PHWR, MW.
The values of a1 and a2 are 0.1935 and 0.1321 t/MW (see
Table S1 in Supplementary Information) respectively (Yue et al.,
2017). Then annual and cumulative natural uranium requirements
can be obtained according to Eq. (1) and Table 1 (the installed
capacity of PWR is assumed to be in a linear increase in each time
period, the same circumstances are also occurred in the other sce-
narios), which are depicted in Fig. 1a and Fig. 1b. The three curves
are corresponding to the low-, medium- and high-speed develop-
ment of installed nuclear power capacity in Table 1.
Annual natural uranium requirements under scenario I are
48,286, 62,798 and 77,311 tons respectively in 2050 for low-,
medium- and high-speed development schemes of China’s
installed nuclear power capacity. And cumulative natural uranium
requirements under scenario I will reach 727,657, 1,155,534 and
1,561,933 tons in 2050 respectively.
2.2.2. Scenario II: The ratio of installed capacity between PWR: PHWR
= 1:0.221 and recycling uranium used in PHWR since the year 2025
(PHWR-RU)
A spent fuel industrial reprocessing plant will be in operation in
the year 2025 in China (China Academy of Engineering, 2011; Hu
et al., 2012; World Nuclear Association, 2015) and recycling ura-
nium is assumed to be used in PHWR since the year 2025 (World
Nuclear Association, 2014). The average U-235 content in spent
fuel of PWR is about 0.80% (Hu et al., 2012). Then spent fuel gener-
ation of 1 MW PWR can support 0.221 MW PHWR (see Table S1 in
Supplementary Information). Assuming the ratio of installed
capacity between PWR and PHWR is 1:0.221, then natural uranium
requirements under this condition will be as the following


M NU ¼ a1  IC PWR þ a2  IC PHWR ð2015—2024Þ
ð2Þ
M NU ¼ a1  IC PWR ð2025—2050Þ
Then annual and cumulative natural uranium requirements can
be obtained according to Eq. (2) and Table 1, which are depicted in
Fig. 2a and Fig. 2b.
 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643 637

Fig. 1. Four potential fuel cycle modes in China’s nuclear power industry in the future NU: Natural Uranium; EU: Enriched Uranium; SF: Spent Fuel; PWR: Pressurised Water
Reactor; PHWR: Pressurised Heavy Water Reactor; SFR: Sodium Fast Reactor; OT: Once Through; RU: Recycling Uranium; MOX: Mixed Oxide; DUPIC: Direct Use of Spent
PWR Fuel in CANDU Reactors; PUREX: Plutonium Uranium Recovery by Extraction.

80000 1600000
Low Low
Medium Medium
60000 High 1200000 High

40000 800000

20000 400000

0 0
2015 2020 2025 2030 2035 2040 2045 2050 2015 2020 2025 2030 2035 2040 2045 2050

Fig. 1a. Annual natural uranium requirements under scenario I (unit: t). Fig. 1b. Cumulative natural uranium requirements under scenario I (unit: t).

Annual natural uranium requirements under scenario II are 969,823(185,711) and 1,307,357(254,576) tons in 2050
39,625(8661, that is comparing with scenario I, natural uranium respectively.
requirements decrease 8661 tons under scenario II), 51,513
(11,286) and 634,00(13,911) tons respectively in 2050 for 2.2.3. Scenario III: The newly-built reactors are all PWRs and mixed
low-, medium- and high-speed development schemes of China’s oxide fuel used in PWR since the year 2025 (PWR-MOX)
installed nuclear power capacity. And cumulative natural uranium MOX fuel used in PWR can increase the uranium fuel use effi-
requirements under scenario II will reach 613,645(114,012), ciency about 1.38 times (China Academy of Engineering, 2011).
638 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643

80000 80000

Low Low
Medium Medium
60000 High 60000 High

40000 40000

20000 20000

0 0
2015 2020 2025 2030 2035 2040 2045 2050 2015 2025 2035 2045

Fig. 2a. Annual natural uranium requirements under scenario II (unit: t). Fig. 3a. Annual natural uranium requirements under scenario III (unit: t)

1600000 1600000

Low Low
Medium Medium
1200000 High 1200000 High

800000 800000

400000 400000

0 0
2015 2020 2025 2030 2035 2040 2045 2050 2015 2020 2025 2030 2035 2040 2045 2050

Fig. 2b. Cumulative natural uranium requirements under scenario II (unit: t). Fig. 3b. Cumulative natural uranium requirements under scenario III (unit: t).

Natural uranium requirements under this condition will be as the
following

8
M NU ¼ a1  IC PWR þ a2  IC PHWR ð2015—2024Þ
ð3Þ
M NU ¼ a3  IC PWRMOX þ a2  IC PHWR ð2025—2050Þ
where a3 is natural uranium requirements per power rating of
PWR-MOX, t/MW, and the value of a3 = 0.1935/1.38 = 0.1405 t/MW.
Then annual and cumulative natural uranium requirements can
be obtained according to Eq. (3) and Table 1, which are depicted in
Fig. 3a and Fig. 3b.
Annual natural uranium requirements under scenario III are
35,113(13,173, that is comparing with scenario I, natural ura-
nium requirements decrease 13,173 tons under scenario III),
45,650(17,148) and 56,188(21,123) tons respectively in 2050
for low-, medium- and high-speed development schemes of Chi-
na’s installed nuclear power capacity. And cumulative natural ura-
nium requirements under scenario III will reach 553,497
(174,160), 872,801(282,733) and 1,174,919(387,014) tons in
2050 respectively.
2.2.4. Scenario IV: SFR put into commercial operation since the year
2035
In the year 2035, the first SFR may be put into commercial oper-
ation in China (China Academy of Engineering, 2011) and is
assumed to account for 1% of total installed nuclear power capac-
ity, account for 5% in 2040, account for 10% in 2045, and account
for 15% in 2050. Besides the SFR, the newly-built reactors are
PWRs. Then natural uranium requirements under this condition
will be as the following
< MNU ¼ a1  IC PWR þ a2  IC PHWR
> ð2015—2024Þ
MNU ¼ a3  IC PWRMOX þ a4  IC PWRPHWR ð2025—2034Þ
>
:
MNU ¼ a3  IC PWRMOX þ a4  IC PWRPHWR þ a5  IC SFR ð2035—2050Þ
ð4Þ
where a4 is natural uranium requirements per power rating of PWR
(NU)-PHWR(RU), t/MW, and the value of a4 = 0.1935/1.221 =
0.1585 t/MW; ICPWR-PHWR means the sum of installed capacity of
PHWR plus the corresponding PWR capacity needed to provided
the recycling uranium for the PHWR, MW; a5 is natural uranium
requirements per power rating of SFR, t/MW, and the value of a5
= 0.0016 t/MW (China Academy of Engineering, 2011); ICSFR is
installed capacity of SFR, MW.
Then annual and cumulative natural uranium requirements can
be obtained according to Eq. (4) and Table 1, which are depicted in
Fig. 4a and Fig. 4b.
Annual natural uranium requirements under scenario IV are
30,061(18,225, that is comparing with scenario I, natural ura-
nium requirements decrease 18,225 tons under scenario IV),
39,036(23,762) and 480,11(29,300) tons respectively in 2050
for low-, medium- and high-speed development schemes of Chi-
na’s installed nuclear power capacity. And cumulative natural ura-
nium requirements under scenario IV will reach 527,351
(200,307), 830,830(324,705) and 1,117,123(444,810) tons in
2050 respectively.
 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643 639

80000 It can be seen from Table 3 that the LCOE for the SFR-MOX is the
highest and the PWR-OT is the cheapest. Cost of investment for the
Low SFR-MOX accounts for the largest proportion and cost of invest-
Meidum ment for the PWR-MOX accounts for the smallest proportion. Fuel
60000 High
related cost for the PWR-MOX accounts for the largest proportion,
whereas for SFR-MOX is negative as the recovered transuranic ele-
ment. Operation and maintenance cost for the SFR-MOX is a little
40000 higher than other fuel cycle modes.
Load factors for these four nuclear fuel cycle modes are all
assumed to be 85%, then the annual cost per unit electricity gener-
20000 ation for them will be as Table 4 shows.
It can be seen from Table 4 that the annual cost of per unit elec-
tricity generation for SFR-MOX is the highest and the value is
0 644.54 (103$/MWa), PWR-OT is the cheapest and the value is
2015 2020 2025 2030 2035 2040 2045 2050 624.05 (103$/MWa).

Fig. 4a. Annual natural uranium requirements under scenario IV (unit: t).
3.2. Optimization model

Three aspects including natural uranium requirements, spent

fuel final disposal amounts and total cost of nuclear power gener-
1600000
ation are considered as the objective functions, the targets are get
the minimum values of them under certain circumstances. Then
Low
Medium the objective functions and constraint conditions are as follows.
1200000 Objective functions:
High
X
2050
MinðRNU Þ ¼ IC i ða1 xi þ a4 yi þ a3 zi þ a5 wi Þ ð5:1Þ
800000 i¼2015

X
2050
MinðF SF Þ ¼ IC i ðb1 xi þ b4 yi þ b3 zi þ b5 wi Þ ð5:2Þ
400000
i¼2015

X
2 050

0 MinðT C Þ ¼ IC i ðm1 xi þ m4 yi þ m3 zi þ m5 wi Þ ð5:3Þ

2015 2020 2025 2030 2035 2040 2045 2050 i¼2015

8
Fig. 4b. Cumulative natural uranium requirements under scenario IV (unit: t).
> IC i ðk1 yi þ k2 zi þ k3 wi Þ 6 RC i
>
>
<x þ y þ z þ w ¼ 1
i i i i
Constraint conditions : ð5:4Þ
2.2.5. Summary >
> y 6 y 6 1
>
:
min i
The cumulative natural uranium requirements under four sce- 0 6 xi ; zi ; wi 6 1
narios analysis are summarized in Table 2.
It can be seen from Table 2 that in the long run, SFR has the where RNU is natural uranium requirements during the period
most advantage in saving natural uranium resources, PWR-MOX 2015–2050, t; FSF is final disposal amount of spent fuel during
comes next, and then PHWR-RU, the last is certainly PWR-OT. 2015–2050, t; Tc is total cost of nuclear power generation during
the period 2015–2050, 103$; ICi is the total installed nuclear power
capacity in the year i, MW; xi, yi, zi, and wi are the ratios of PWR-OT,
3. Optimization model for nuclear fuel cycles in China PHWR-RU, PWR-MOX, and SFR-MOX installed capacity accounted
for the total installed nuclear power capacity in the year i,%; a1,
3.1. Data preparations a4, a3, and a5 are natural uranium requirements per power rating
of PWR, PWR(NU)-PHWR(RU), PWR-MOX, and SFR, t/MW; b1, b4,
In order to compare with the cost for the different fuel cycle b3, and b5 are spent fuel final disposal amount per unit power rating
modes, levelized cost of electricity (LCOE) is used here. LCOE is for PWR-OT, PHWR-RU, PWR-MOX, and SFR-MOX, t/MW (see Sup-
the total cost under a certain fuel cycle mode included, and divide plementary Information S1); m1, m4, m3, and m5 are annual cost per
it into the electricity generation cycle in the service period. The unit electricity generation for PWR-OT, PHWR-RU, PWR-MOX and
LCOE under different nuclear fuel cycle modes is depicted in Table 3 SFR-MOX, 103$/MWa (see Table 4); k1, k2, and k3 are reprocessing
(Ding, 2014; Massachusetts Institute of Technology, 2003, 2009). capacity demand coefficient per unit power rating for PHWR-RU,

Table 2
Cumulative natural uranium requirements under the four scenarios in the year 2020, 2030 and 2050 (unit: t).

2020 2030 2050

Low Medium High Low Medium High Low Medium High
Scenario I 44092 57250 66925 176306 263256 334996 727657 1155534 1561933
Scenario II 44092 57250 66925 160557 237474 300882 613645 969823 1307357
Scenario III 44092 57250 66925 152108 223865 282989 553497 872801 1174919
Scenario IV 44092 57250 66925 153051 224807 283931 527351 830830 1117123
640 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643

Table 3
Levelized cost of electricity under different nuclear fuel cycle modes (Unit: 103$/KWh).

Cost of investment Fuel related cost Operation and maintenance cost Levelized cost of electricity
PWR-OT 67.68 8.40 7.72 83.81
80.76% 10.03% 9.22% 100%
PHWR-RU 68.08 8.15 7.74 83.97
81.08% 9.71% 9.21% 100%
PWR-MOX 67.68 9.97 7.72 85.38
79.27% 11.68% 9.05% 100%
SFR-MOX 81.23 -3.911 9.25 86.57
93.83% -4.52% 10.69% 100%

Note: 1. Fuel related cost for SFR-MOX is negative as the fuel made from the recovered transuranic element can obtain high cost deduction.

Table 4 Then kmax can be calculated as kmax = 3.065.

Annual cost of per unit electricity generation for PWR-OT, PHWR-RU, PWR-MOX and
SFR-MOX (Unit: 103$/MWa).
 Consistency check
PWR-OT PHWR-RU PWR-MOX SFR-MOX
624.05 630.47 635.72 644.54 Based on the formula, CR = CI/RI, where CR is the consistency
ratio, CI is the consistency index, and RI is the randomness index.
By the corresponding formula CI = (kmax  n)/(n  1) and values of
PWR-MOX, and SFR-MOX, t/MW (see Supplementary Information kmax = 3.065, n = 3, RI = 0.58. Then CR can be calculated as CR =
S1); RCi is reprocessing capacity of spent fuel in the year i, t (see 0.056 < 0.1, so the judgment matrix (P) has the satisfactory
Supplementary Information S2); ymin is the existed PHWR ratio consistency.
(the PHWR installed capacity is 1456 MW at present).

3.3. Model resolved 4. Results and discussions

Three objection functions are included in the optimization
model, so it belongs to multi-objective optimization. The weight-
ing method is applied here to transform the multi-objective prob-
lem into single objective problem, and then MATLAB is used to
solve the model.
The determination of weighting value includes the following
three steps (Chang and Jiang, 2007).
4.1. Optimization results
Based on the eigenvectors obtained in Section 3.3 and by the
MATLAB software, the optimization results under scenario A and
scenario B of reprocessing capacity (see Supplementary Informa-
tion S2) can be solved.

(1) constructing judgment matrix (P) (Table 5) 4.1.1. Results under scenario A of reprocessing capacity
For the high speed of installed nuclear power capacity in China
According to the meaning of 1  9 in AHP method (Chang and described in Table 1, the optimization results under scenario A of
Jiang, 2007), the weighting values of the three indicators ZC, ZSF, reprocessing capacity are depicted in Fig. 5.
and ZNU are determined. In the period 2015–2024, as China doesn’t built commercial
spent fuel reprocessing plant in this period, the reactors are ‘‘once
(2) Solving the eigenvectors (W) through” fuel cycle mode. At the same time, PHWR will not be
built during this period and PWR-OT proportion (Xi) will rise year
According to Table 5, the eigenvectors W can be calculated as by year. In the period 2025–2034, the proportion of recovered
W=(W1, W2, W3)=(0.072, 0.279, 0.649). uranium used in PHWR (Yi) decline year by year, which is mainly
due to limited spent fuel reprocessing capacity in China. The scale
(3) Consistency check of reprocessing capacity development speed is far behind the
expansion of installed nuclear power capacity, leading PWR-OT
 Calculate the maximum eigenvalue of judgment matrix (kmax) to increase year by year. In the year 2035 and 2045, the propor-
tion of recovered uranium used in PHWR has an increase as two
1X 3
ðPWÞi 800 t/a spent fuel reprocessing plants are put into commercial
kmax ¼ ð6Þ
n i¼1 W i operation in these two years. The trends during the period
2035–2044 and 2045–2050 are similar to the period of 2025–
where (PW)i is the element i of PW, n is 3 here. 2034, the proportion of recovered uranium used in PHWR will
decline year by year.
2 3 2 32 3 2 3
ðPWÞ1 1 1=5 1=7 0:072 0:2205
6 7 6 76 7 6 7
and PW ¼ 4 ðPWÞ2 5 ¼ 4 5 1 1=3 54 0:279 5 ¼ 4 0:8553 5
4.1.2. Results under scenario B of reprocessing capacity
ðPWÞ3 7 3 1 0:649 1:990 For the high speed of installed nuclear power capacity in China
described in Table 1, the optimization results under scenario B of
Table 5
reprocessing capacity are depicted in Fig. 6.
Judgment matrix (P).
Comparing Figs. 5 and 6, it can be seen that with the expansion
P ZC ZSF ZNU of spent fuel reprocessing, the proportions of PHWR-RU and SFR-
ZC 1 1/5 1/7 MOX are both relatively increased. So the most important factor
ZSF 5 1 1/3 restricting the adjustment of fuel cycle mode’s structure is the
ZNU 7 3 1
scale of spent fuel reprocessing ability.
 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643 641

1
0.9 Xi Yi Zi Wi
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2015 2020 2025 2030 2035 2040 2045 2050

Fig. 5. Optimization results under scenario A of reprocessing capacity.

0.9 Xi Yi Zi Wi
0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
2015 2020 2025 2030 2035 2040 2045 2050

Fig. 6. Optimization results under scenario B of reprocessing capacity.

4.2. Discussions
4.2.1. Annual and cumulative natural uranium requirements under
scenario A of reprocessing capacity
For the high speed of installed nuclear power capacity in China
described in Table 1, the annual natural uranium requirements
under scenario A of reprocessing capacity can be calculated accord-
ing to Eq. (7).
to the year 2050 (compared with before optimization), accounts
for 5.19% of the total natural uranium demand before optimization.
By the same method for the optimization of low speed and
medium speed of installed nuclear power capacity (Table 1), the
cumulative savings of natural uranium requirements to the year
2050 (compared with before optimization) are 81,724 t and
80,514 t, respectively, which are accounted for 11.34% and 7.01%
of the total natural uranium demand before optimization.

ARNU ¼ a1 xi þ a4 yi þ a3 zi þ a5 wi
where ARNU is annual requirements of natural uranium, t/a; the def-
initions and values of a1, a4, a3, a5 and xi, yi, zi, wi have been given in
Section 2.2 and Section 4.1.1 respectively. So the corresponding
annual and cumulative natural uranium requirements under sce-
nario A can be calculated and are depicted in Fig. 7.
Natural uranium requirements start to reduce in the year 2025,
and repeated reduce in the year 2035 and 2045 as reprocessing
capacity are enlarged in these two years and SFR is put into use
in 2035. But the decrement is not large as the relatively small scale
of reprocessing capacity.
After the optimization, the cumulative natural uranium require-
ments start to decline in the year 2025 comparing with wholly
PWR-OT fuel cycle, and the decline increases gradually. But due
to the limitation of spent fuel reprocessing capacity in China, the
cumulative savings of natural uranium resources is only 80,652 t
ð7Þ
4.2.2. Annual and cumulative natural uranium requirements under
scenario B of reprocessing capacity
For the high speed of installed nuclear power capacity in China
described in Table 1, the annual natural uranium requirements
under scenario B of reprocessing capacity can be calculated accord-
ing to Eq. (7).
The values of a1, a4, a3, a5 and xi, yi, zi, wi have been given in Sec-
tion 2.2 and Section 4.1.2 respectively. So the corresponding
annual and cumulative natural uranium requirements under sce-
nario B can be calculated and are depicted in Fig. 8.
Natural uranium requirements start to reduce in the year 2025,
and repeated reduce in the year 2035 and 2045 as reprocessing
capacity are enlarged in these two years and SFR is put into use
in 2035. And the decrement is bigger than that in scenario A of
reprocessing capacity as the relatively larger scale of reprocessing
capacity.
642 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643

80000 1600000

Annual natural uranium requirements

Before opmizaon(annual)
70000
Aer opmizaon(annual)

Cumulave natural uranium

60000 Before opmizaon(cumulave) 1200000
Aer opmizaon(cumulave)
50000

requirements
40000 800000

30000

20000 400000

10000

0 0
2015 2020 2025 2030 2035 2040 2045 2050

Fig. 7. Annual and cumulative natural uranium requirements before and after optimization under scenario A (unit: t).

80000 1600000
Before opmizaon(annual)
Annual natural uranium requirements

70000 Aer opmizaon(annual)

Cumulave natural uranium

Before opmizaon(cumulave)
60000 1200000
Aer opmizaon(cumulave)

requirements
50000

40000 800000

30000

20000 400000

10000

0 0
2015 2020 2025 2030 2035 2040 2045 2050

Fig. 8. Annual and cumulative natural uranium requirements before and after optimization under scenario B (unit: t).

After the optimization, the cumulative natural uranium require-
ments start to decline in the year 2025 comparing with wholly
PWR-OT fuel cycle, and the decline increases gradually. Due to
the limitation of spent fuel reprocessing capacity in China, the
cumulative savings of natural uranium resources is 139,315 t to
the year 2050 (compared with before optimization), accounts for
8.96% of the total natural uranium demand before optimization.
By the same method for the optimization of low speed and
medium speed of installed nuclear power capacity (Table 1), the
cumulative savings of natural uranium requirements to the year
2050 (compared with before optimization) are 124,065 t and
138,742 t, respectively, which are accounted for 17.21% and
12.08% of the total natural uranium demand before optimization.
In order to maximize the saving of natural uranium resource
consumption and the reduction of spent fuel final disposal
amounts, research and investment on the advanced fuel cycle
modes should be enhanced, as well as the expansion of the scale
for spent fuel reprocessing capacity so as to increase the uranium
use efficiency. As these are two of the most critical factors that
restricting the sustainable development of the nuclear power
industry in China at present and in the future.
uranium resources and reduces the amount of spent fuel disposal.
Fast reactor fuel cycle has obvious advantage of saving uranium
resources.
Then taking the natural uranium requirements, spent fuel final
disposal amounts and total cost of electricity generation as the
optimized objectives, restricted by the different reprocessing
capacity, structure of fuel cycle modes in nuclear power industry
is optimized in 2015–2050. The results show that the mode of
PHWR-RU is more competitive when spent fuel reprocessing
capacity is small. The uranium resource saving rate and the reduc-
tion rate of spent fuel disposal (see Supplementary Information S3)
increase with the reprocessing capacity increases. So it is the scale
of reprocessing capacity that is most critical for restricting the
structure regulation of fuel cycle mode of China’s nuclear power
industry in the future.
Acknowledgements
This research was supported by the National Natural Science
Foundation of China (71373003).
Appendix A. Supplementary data

5. Conclusions
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.anucene.2017.09.
The natural uranium requirements are predicted under three
049.
development patterns (low-, medium-, and high-speed) based on
the national nuclear power development’s plan of China’s nuclear
References
power industry till the year 2050 with reprocessing fuel cycle
mode and ‘‘once-through” fuel cycle mode. Compared to ‘‘once- Bernard, P., 2007. French experiences and perspectives on plutonium recycling in
through” fuel cycle mode, reprocessing fuel cycle mode conserves the existing power fleet. Prog. Nucl. Energy 49, 583–588.
 Q. Yue et al. / Annals of Nuclear Energy 111 (2018) 635–643
Cao, B., Wu, J., Yang, T.R., Ma, X.B., Chen, Y.X., 2013. Preliminary study on nuclear
fuel cycle scenarios of China before 2050. Energy Procedia 39, 294–299.
Chang, J.E., Jiang, T.L., 2007. Research on the weight of coefficient through analytic
hierarchy process. J. Wuhan Univ. Technol. (Inf. Manage. Eng.) 29 (1), 153–156
(in Chinese).
China Academy of Engineering, 2011. Strategic Study on China Energy Long Term
(2030, 2050) Development. Science Press, Beijing (in Chinese).
Ding, X.M., 2014. Economic analysis of fast reactor fuel cycle with different modes.
China Nucl. Power 7 (2), 160–167 (in Chinese).
Hu, P., Zhao, F.Y., Wang, Z., 2012. The contrastive study of spent fuel characteristic
in different PWR nuclear fuel cycles. Sustainable Energy 2, 51–55 (in Chinese).
Kim, P.O., Lee, K.J., Lee, B.W., 1999. Selection of an optimal nuclear fuel cycle
scenario by goal programming and the analytic hierarchy process. Ann. Nucl.
Energy 26 (5), 449–460.
Kunsch, P.L., Teghem, J., 1987. Nuclear fuel cycle optimization using multi-objective
stochastic linear programming. Eur. J. Oper. Res. 31 (2), 240–249.
Liu, X.G., Xu, J.M., Zhu, Y.J., 2006. Chinese nuclear power development and related
fuel cycle scenarios. Sci. Technol. Rev. 24 (6), 22–25 (in Chinese).
Ma, X.B., Chen, Y.X., Wang, L.Z., Cao, B., Lu, D.G., 2013. Preliminary study on PWR
nuclear fuel cycle scenarios. At. Energy Sci. Technol. 47 (5), 811–815 (in
Chinese).
Massachusetts Institute of Technology, 2003. The future of nuclear power. 2003.
http://web.mit.edu/nuclearpower/.
Massachusetts Institute of Technology, 2009. Update of the MIT 2003: The future of
nuclear power. 2009. http://web.mit.edu/nuclearpower/.
National Development and Reform Commission, 2007. National Nuclear
Long-and-medium Term Development Planning (2005-2020). http://
www.gov.cn/gzdt/att/att/site1/20071104/00123f3c4787089759a901.pdf.2007.
(in Chinese).
643
Nuclear Energy Agency, 2015. International Energy Agency. Technology roadmap:
Nuclear.
OECD Nuclear Energy Agency, 2014. International Atomic Energy Agency. Uranium
2014: Resources, Production and Demand.
Park, B.H., 2017. Assessment of spent nuclear fuel amounts to be managed based on
disposal option in Republic of Korea. Ann. Nucl. Energy 109, 199–207.
World Nuclear Association. The AFCR and China’s fuel cycle. http://www.world-
nuclear-news.org/E-The-AFCR-and-Chinas-fuel-cycle-11111401.html, 11
November 2014, 2014.
World Nuclear Association. China’s Nuclear Fuel Cycle. http://www.world-nuclear.
org/info/Country-Profiles/Countries-A-F/China-Nuclear-Fuel-Cycle/, Updated
April 2015.
Worrall, A. 2013 Utilization of used nuclear fuel in a potential future US fuel cycle
scenario. In: WM2013 Conference, February 24–28, 2013, Phoenix, Arizona USA.
Yolanda Moratilla Soria, B., Ruiz-Sánchez, R., Estadieu, M., Belda-Sánchez, B.,
Cordón-Peralta, C., Martín-Cañas, P, Rodriguez-Penalonga, L., del Mar Cledera-
Castro, M., Ana Sáenz-Nuño, M., Morales-Polo, C., 2015. Impact of the taxes on
used nuclear fuel on the fuel cycle economics in Spain. Energies 8 (2), 1426–
1439.
Yoon, S., Choi, S., Ko, W., 2017. An integrated multicriteria decision-making
approach for evaluating nuclear fuel cycle systems for long-term sustainability
on the basis of an equilibrium model: Technique for order of preference by
similarity to ideal solution, preference ranking organization method for
enrichment evaluation, and multiattribute utility theory combined with
analytic hierarchy process. Nucl. Eng. Technol. 49, 148–164.
Yue, Q., He, J.K., Stamford, L., Azapagic, A., 2017. Nuclear power in China: an analysis
of the current and near-term future uranium flows. Energy Technol. 5, 681–691.
Zhang, J.P., Liu, Z.R., Wang, L., 2016. Uranium demand and economic analysis of
different nuclear fuel cycles in China. Energy Strategy Rev. 9, 50–61.