Nuclear Power in China: An Analysis of the Current and Near-Future Uranium Flows



Nuclear power capacity is growing rapidly in China and is expected to reach 58 GW by 2020. However, it is not clear at present if the country can meet this increasing demand for uranium. We present a comprehensive uranium flow analysis in China that covers a ten-year period from 2003 to 2012 with the aim of identifying near-future fuel demand to 2020. The findings suggest that from 2003 to 2012, 19 126 t of natural uranium were consumed, which generated 3811 t of spent fuel. Of this, only 33–58 % of natural uranium was sourced domestically, the rest of which was imported. By 2020, the uranium consumption is estimated to reach 14 426 t yr−1, around three quarters of the amount consumed over the whole ten-year period considered here. Based on the current increase in demand, the domestic natural uranium could run out by 2027 if the country relies solely on its own uranium deposits known currently. Our findings also show that China would need to expand its spent-fuel reprocessing program, expected to start in 2025, by 3–4 times as the planned capacity is insufficient given the current growth of nuclear power.


China's ongoing economic development has led to an ever-growing energy demand, which is likely to continue for some time.[1] A particular increase is expected in electricity consumption because of the rapid growth of China's tertiary industry, which relies heavily on electricity.[2] The total installed electrical capacity is expected to grow by nearly 70 % between 2020 and 2050 (Table 1). To meet this substantial increase in demand and to attempt to curb greenhouse gas emissions, nuclear and other low-carbon power would have to play a greater role in China's electricity mix, with the former expected to contribute 15–24 % to the total electricity generation by 2050 (Table 1). By comparison, in 2012, nuclear power generated only 2 % of the total electricity.[3] Therefore, sustainable supply and use of nuclear fuel as well as management of spent fuel are critical issues for nuclear power in China.

Table 1. Projected installed capacity of electricity plants in China for the period 2020–2050 and the expected contribution from nuclear power.[2, 4]
Installed capacity and power generation202020302050
total installed capacity of electricity plants [GW]150020002500
installed capacity of nuclear power plants [GW]5883.8–200250–400
ratio of installed nuclear power capacity to the total installed electricity capacity [%]3.84.2–1010–16
ratio of nuclear power generation to the total power generation [%]5.86.3–1515–24

However, the deployment of nuclear power at the rate anticipated in China is challenging. The projected installed capacity of 250–400 GW by 2050 (Table 1) is comparable to today's global installed capacity of operating reactors of 384 GW.[5] Consequently, assuming the same technologies and fuel cycles, this implies a doubling of global annual uranium demand from today's 65 000 tU,[5] even without considering further nuclear deployment outside of China. In reality, technologies and fuel cycles will change from the current global average and, therefore, it is not possible to predict the actual uranium demand implied by the 2050 figures above. For example, sodium fast reactors are expected to be brought online in the 2030s in China, which would help to reduce the pressure on uranium resources.[2] Nevertheless, it is clear that China's currently available uranium resources are insufficient to meet the growing demand. Uranium ores are predominantly low grade and contain 0.05–0.3 % of uranium at best. At the same time, more than 60 % of China's total reserves are small- and medium-sized ore bodies.[2, 6] In the period from 2003–2012 the uranium production in China ranged between 710 and 1450 tU.[7, 8] Therefore, it is not surprising that imports of uranium have been increasing[9] and will continue to do so unless China finds ways for a more sustainable use of uranium resources, including the use of spent fuel.

Uranium utilization and spent fuel have been the subjects of considerable work, mainly for other world regions. The previous studies include, for instance, fuel cycle options for Europe[10] and Korea,[11] whereas spent fuel has been studied for the USA[12] and Europe.[13] Other authors focused on the evaluation of different fuel cycles. For example, Gao and Ko studied material flows for 13 different fuel cycle options using an equilibrium model[14] and assessed their sustainability based on five criteria: resource reliability, environmental impacts, proliferation resistance, economic competitiveness, and technological readiness.[15] A focus on environmental impacts in particular has motivated the evaluation of different fuel cycles relative to a conventional open cycle (in which no reprocessing occurs and spent fuel is treated as waste) both in general using life cycle assessment[16] and in the French context using ecological footprinting.[17] Life cycle assessment has also been used to consider the sustainability implications of nuclear power in the UK.[18]

However, few studies have been carried out for fuel utilization in China's nuclear power industry. Previous work includes an analysis of pre-2020 scenarios for natural uranium demand and separative work as well as the amount of spent fuel arising.[19] Similar research has also considered potential scenarios to 2050.[20] No previous work has attempted to analyze the flows and stocks of nuclear fuel in China, so that at present, it is not clear if and how the country would be able to meet the fast-growing demand for this resource and what implications that may have globally in a resource-constrained world. In an attempt to provide an answer to this question, we present a comprehensive fuel flow analysis for China's nuclear power industry in the period from 2003 to 2020 with the aim of estimating current and near-future uranium demand and spent fuel generation. As far as we are aware, this is the first such analysis for China.


In this section, we first outline the uranium fuel cycle that has been followed to estimate the uranium flows in China. The estimates are based on the method developed in this work, which is presented in the subsequent section.

An overview of the uranium life cycle

The nuclear fuel life cycle involves the following stages: uranium extraction, conversion and enrichment, fuel fabrication, electricity generation, and waste management. These are outlined in Figure 1 and described in turn below. The uranium flows shown in Figure 1 were estimated using the equations and data described in the section on the estimation of uranium flows in China.

Figure 1.

Flows and stocks of uranium in China in 2012 in different stages of the nuclear fuel cycle [t yr−1] (excluding uranium fuel demand for the first loading of new reactors). The dashed line represents flows of recycled uranium, which are currently zero as there is no fuel reprocessing in China. Only spent fuel from PWR is shown as potentially recyclable because it contains a higher content of 235U and 239Pu than PHWR spent fuel.

Uranium extraction

Uranium can be extracted from the ground using open-pit or underground mining or in situ leaching. The former two are followed by milling to extract the uranium, whereas the latter involves the use of leaching agents to dissolve it from the ore underground; the solution is then pumped to the plant in which uranium is recovered by using an ion-exchange system. Mining used to be the most common way to extract uranium but, in recent years, in situ leaching has become more widespread and now accounts for nearly half of the world's uranium mining.[21] Regardless of the extraction technique, the product of this stage is in the form of uranium oxide (U3O8), known as “yellow cake”.


The yellow cake that leaves the uranium-recovery plant must be enriched before it can be used as reactor fuel. The exceptions to this are reactor types that do not require enriched uranium, for example, pressurized heavy-water reactors (PHWR), in which case the U3O8 is converted directly to UO2. The enrichment process requires the materials to be in gaseous form, so the yellow cake must be converted into uranium hexafluoride (UF6).


Natural uranium contains only 0.7 % of the fissile 235U isotope and virtually all of the remaining 99.3 % is non-fissile 238U. Fuel enrichment involves increasing the proportion of 235U from 0.7 % to approximately 3–4 %, depending on the type of reactor used. Centrifugation is currently the main commercial method used to accomplish this step. During this stage, most of the uranium content is lost and stored as depleted uranium (DU), which has a 235U content of approximately 0.2–0.3 %.

Fuel fabrication

The enriched UF6 is transported to a fuel-fabrication plant to be converted into UO2 powder. The powder is baked and pressed into pellets that are stacked and encased into thin alloy or steel tubes (the cladding) to produce fuel rods. A number of fuel rods are sealed together to produce “fuel assemblies”, which are then inserted into the nuclear reactor.

Electricity generation

As the nuclear fission reaction proceeds in the reactor core, the amount of 235U decreases and the amount of fission products, such as 137Cs, increases. Some of the 238U content can also be converted to other products, such as actinides, including 239Pu.

Waste management

The great majority of the radioactivity produced in the nuclear life cycle is contained in the spent fuel. For example, spent fuel from a pressurized water reactor (PWR) contains up to 96 % 238U, 3 % radioactive fission products, and 1 % each of 235U and 239Pu.[22] Initially, spent fuel is kept in an interim storage in a pool of water for several months to a few years to allow the radiation levels to decrease. After that, it can be recycled to recover uranium and plutonium for reuse as nuclear fuel or it can be stored until it can be managed safely. Spent fuel from a PWR is more readily (and economically) recyclable as it has a higher content of 235U and 239Pu than that from PHWR. In the most commonly used process, plutonium uranium recovery by extraction (PUREX), this is performed by dissolving the fuel and the cladding in an acid to separate the components. Uranium can be returned to the conversion plant to be further enriched to 3–4 % 235U or combined with the plutonium to produce mixed oxide (MOX) fuel. The reprocessing stage can also produce transuranium elements such as californium. However, most of the radioactive waste arising in the nuclear life cycle is stored. The type of storage depends on the level of radioactivity in the waste, which in turn is used to classify it into the following categories: high-level waste (HLW), intermediate-level waste, and low-level waste.

Estimation of uranium flows in China

By the end of 2012, there were 15 nuclear reactors in commercial operation in China, 13 of which were PWRs and two were PHWRs. Their total installed capacity reached 12.6 GWe (Table 2), equivalent to 1.1 % of China's total power capacity. They generated 98.3 TWh or 1.97 % of total electricity generation, an increase of 12.8 % on the previous year. A breakdown of the generation by power plant is given in Table S1. In the same year, 30 new nuclear plants were under construction in mainland China, with a total installed capacity of 32.7 GW (Table 2 and Table S2). This is nearly half the number (66) and capacity (65.5 GW) of nuclear plants currently under construction in the rest of the world.[3, 23] Typically, it takes around five years for a new power plant to be built in China.

Table 2. Nuclear power in China over the period 2003–2012.[3]
Installed capacity and generation2003200420052006200720082009201020112012
installed capacity [GWe]6.356.966.966.969.089.089.0810.2111.9412.63
under construction [GWe]2.772.123.204.924.5411.0221.9230.9729.2432.67
power generation [TWh]43.850.553.154.862.969.270.174.787.298.3

At present, spent fuel is not recycled in China. However, the government is planning to introduce fuel reprocessing in the future and, as a first step, a pilot plant was built in 2010 at Lanzhou Nuclear Fuel Complex in Gansu province.[2] Using the PUREX process, the plant was designed to reprocess 50 t of used fuel over the period 2013–2015. A commercial reprocessing plant based on indigenous advanced technology with a capacity of 800–1000 t yr−1 was planned to begin operation by 2025; consequently, the recycling and use of post-processed uranium and plutonium will be very limited before then,[2, 24, 25] and nuclear fuel will continue to be produced from natural uranium.

To determine the amount of natural uranium and other related flows in the fuel life cycle in China, the following parameters were estimated in this work:

  1. the amount and efficiency of fuel use;

  2. the requirements for natural uranium (235U) and the level of self-sufficiency;

  3. the requirements for fuel conversion and enrichment as well as the related generation of depleted uranium;

  4. the separative work unit (SWU) requirement; and

  5. the amount and efficiency of spent fuel generation.

These are discussed in turn below.

i) Uranium fuel consumption MFuel [tU] has been estimated based on the thermal power of the reactors, their load factor and burn-up rate as follows [Eq. (1)]:[19, 20]

display math(1)

where NT is the installed thermal power [MW], 365 is the number of days in a year, CF is the load factor, and Bd is an average discharge burn-up [MWd tU−1]. NT and Bd are given in Table 3 for each reactor and the load factors over the period are detailed in Table S3.

Table 3. Estimated uranium flows in China in 2012 based on the operating conditions of nuclear power plants.
Nuclear power plantReactor no.Type[a]Installed capacity [MWe][a]NT [MW][b]P [GWh yr−1]Bd[b] [GWd tU−1]CF [%][a]Enrichment level[b] [%]MFuel [Eq. (1)] [tU]MNat [Eq. (3)] [tU]S [Eq. (8)] [tSW]MSF [Eq. (9)] [t]
  1. [a] Source: Ref. [3]. [b] Enrichment level of fuels; sources for PWR: Refs. [24, 27, 28]; sources for PHWR: Refs. [8, 28, 29].

Qinshan Iunit onePWR310966284424101.23.014.989.957.214.6
Qianshan IIunit onePWR650193020 1623584.73.2517.0112.374.116.8
unit twoPWR65019303581.13.2516.3107.570.916.0
unit threePWR65019303590.73.2518.3120.379.317.9
unit fourPWR65019303596.83.2519.5128.484.719.2
Qinshan IIIunit onePHWR728206111 6277.1597.4102.5102.598.9
unit twoPHWR72820617.1591.796.496.493.0
Dayawanunit onePWR984290515 9303383.93.227.0174.7114.426.5
unit twoPWR984290533100.53.232.3209.2137.031.7
Lingaounit onePWR990290531 5133391.93.729.5223.8155.429.0
unit twoPWR99029053389.73.728.8218.5151.728.3
unit threePWR108029053386.33.727.7210.2146.027.3
unit fourPWR108029053378.53.725.2191.3132.824.8
Tianwanunit onePWR1060300016 2414386.73.922.1177.1125.321.7
unit twoPWR106030004387.73.922.3179.2126.822.0

The efficiency of nuclear fuel use EFuel [kWh gU−1] relative to the annual generation of power P is estimated as [Eq. (2)]:

display math(2)

ii) The estimated amount of uranium fuel was then used to calculate the requirement for natural uranium Mnat (0.7 % 235U; [tU]) for each reactor over the period according to Equation (3):[19, 20]

display math(3)

where MConv is the input of UF6 in the enrichment stage [tU], r is the uranium recovery rate during natural uranium conversion, production of UO2, and fuel fabrication, with r equal to 0.995 for each step.[19, 20] MEn is the amount of enriched uranium [tU], XNat, XEn, and XDU represent the percentage content of 235U in the natural, enriched, and depleted uranium, respectively, with XNat=0.7 %, XEn=3–4 %, and XDU=0.2–0.3 %.[19, 20]

Natural uranium comes from two sources: domestic production and imports from overseas markets. Many factors affect the uranium mining sector, including techno-economic, environmental, and social. However, domestic production might be seen as more secure than imports as the latter are affected more easily by external factors, including international trade agreements and the activities of foreign companies in their home markets. Therefore, it is important to determine the level of China's self-sufficiency with respect to natural uranium, particularly as the pressure on this resource is expected to grow rapidly in the future. For these purposes, a self-sufficiency index (SSI [%]) is proposed [Eq. (4)]:

display math(4)

where MNat (Dom) represents the amount of natural uranium produced domestically over a certain period and MLoad is the amount of natural uranium at first loading of new reactors.

iii) Based on the amount of natural uranium estimated by using Equation (3), the amount of UF6 [tU] that can be obtained in the conversion process can be estimated according to Equation (5):

display math(5)

where LConv is the rate of uranium loss in the conversion process, which is equal to 0.005.[19, 20]

The amount of enriched uranium MEn (3–4 % 235U; [tU]) can be calculated as [Eq. (6)]:

display math(6)

where LFabr represents the loss rates in the production of UO2 and the reactor fuel, both equal to 0.005, as in the conversion process.[19, 20] The amount of depleted uranium MDU [tU] is then estimated as the difference between the converted and enriched uranium [Eq. (7)]:

display math(7)

As PHWRs do not require uranium enrichment, only the direct conversion of U3O8 to UO2 is needed and hence the uranium loss rate is very low and can be neglected. Thus, uranium fuel consumption for PHWRs is approximately equal to the natural uranium consumption.

iv) Separative work unit (SWU) requirement:

Using the parameters in Equations (1) and (2), the SWU requirement S [tSW] can be determined as [Eq. (8)]:[19]

display math(8)

v) Amount and efficiency of spent fuel generation:

The amount of spent fuel MSF [t] generated during the operation of a nuclear reactor over a certain period can be estimated as [Eq. (9)]:

display math(9)

where LFuel is the loss rate of uranium fuel through burn-up during the operation of the reactor. In PWRs, the loss rate is equal to 0.0168,[26] and in PHWRs it is 0.0348.[11] As an illustration, the amount of spent fuel generated by each reactor in 2012 is given in Table 3.

The amount of each component Mi [tU, tPu, or tFP (FP=fission products)] in the spent fuel can be calculated as [Eq. (10)]:

display math(10)

where αi represents the fraction of each component.

To determine the reactor efficiency ESF [kWh g] in terms of the amount of spent fuel generated relative to the power generated P, Equation (11) can be used:

display math(11)

Results and Discussion

The results obtained using the above equations are first presented and discussed for the ten-year period from 2003 to 2012, followed by an analysis of the expected near-future natural uranium demand and spent fuel generation up to 2020. Note that beyond 2012 detailed data were not available for nuclear power in China, so that the uranium flows up to 2020 are estimated using the average parameters for the 2003–2012 period and the installed (2013–2016) or projected (2017–2020) capacities of nuclear plants.

Uranium flows from 2003–2012

As an illustration of how the uranium flows were estimated for each reactor and year, the results for the year 2012 are detailed in Table 3 and the summary for the whole year is given in Figure 1. The same calculations were performed for all other years within the considered period to estimate the total amount of uranium fuel consumed in China from 2003–2012.

The total amount of fuel consumed by China's nuclear power plants in 2012 was around 500 tU (Table 3 and Figure 1). This is equivalent to the total natural uranium consumption of 2341 tU (235U) with the SWU requirement of 1455 tSW (1.45 m SWU). Of the total natural uranium demand, 1450 tU was produced domestically[7] and 891 tU was imported from overseas and used in 2012. Note that the total amount of imported uranium in that year was much higher (12 908 tU);[25] it is unclear what happened to the rest but it is presumably stored for future use. A total of 1828 tU of DU was produced in the enrichment process and stored. Around 488 tU of spent fuel was generated, which was also stored.

The corresponding results for the whole period from 2003–2012 are summarized in Table 4 and Figure 2. As can be seen, natural uranium consumption used in the fuel cycle more than doubled from 1133 to 2341 tU during the period, excluding the fuel used for the first loading of new reactors.

Table 4. Estimated flows and stocks of uranium in China from 2003–2012 in different stages of the fuel cycle.
YearMNat (Dom)[a] [tU]Imported MNat [tU]Total MNat requirement[b] [tU]MNat consumption[c] [tU]MConv[d] [tU]MEn[e] [tU]MDU [tU]MFuel[f] [tU]MSF [t]
  1. [a] U3O8; all data from Ref. [7], except for the years 2009 and 2010, which are from Ref. [8]. [b] Total natural uranium requirement includes natural uranium consumption in the fuel cycle and natural uranium demand on first loading of new reactors; for example, for a new 1000 MW reactor, 405 tU is required on first loading.[2] [c] MNat as estimated by Eq. (3); excludes natural uranium used for the first loading of new reactors. [d] UF6 (0.71 % 235U). [e] UF6 (3–4 % 235U). [f] UO2.

total886710 25919 12615 89015 820200811 89039113811
Figure 2.

Consumption of natural uranium and spent fuel generation based on the installed nuclear plant capacity in the period 2003–2012.

Overall, the total uranium consumption over the period was 19 126 tU, which included 3236 tU used for the first loading of new reactors (Table 4). Compared with natural uranium consumption, the rate of spent fuel generation was slower, increasing by 53 %, from 318 t at the beginning of the period to 488 t by 2012, with the total amount of spent fuel reaching 3811 t over 10 years. The slower rate of increase in spent fuel generation compared to natural uranium consumption can be attributed to the increasing uranium fuel use and spent fuel efficiencies (Figure 3). The efficiencies increased in three distinct steps over the period: in 2004, 2007, and 2011. This is because of new, more efficient units that came online in these three years (Table 5).

Figure 3.

Uranium self-sufficiency index (SSI), fuel efficiency (EFuel), and spent fuel generation efficiency (ESF) over the period 2003–2012.

Table 5. Fuel consumption and other parameters for PWRs in China over the period 2003–2012.
installed capacity [MW]4908555855585558767876787678875810 48811 138
power generation [GWh yr−1]35 91639 96642 97643 34251 35957 96258 37763 28875 69986 673
uranium fuel consumption [tU yr−1]137.3136.8161.9158.7155.6206.2207.7241.3281.3300.9
uranium fuel consumption per unit of installed capacity [kgU MW−1]27.627.529.128.628.026.927.127.626.827.027.6
natural uranium consumption [tU yr−1]943.1941.11108.6109010671458.714711728.520162142
natural uranium consumption per unit of installed capacity [kgU MW−1]192192200196192190192197192192194
separative work unit [tSW yr−1]632.2631.2730.5714.7742.2988.8997.911771372.91455.4
separative work unit per unit of installed capacity [kg SW MW−1]129129131129134129130134131131131
spent fuel generation [t yr−1]135134.5159.1156153202.7204.2237.2276.6295.8
spent fuel generation per unit of installed capacity [kg MW−1]27.527.428.628.127.526.426.627.126.426.627.2
plutonium content in spent fuel per unit of installed capacity [gPu MW−1]275274286281275264266271264266272
fission products in spent fuel per unit of installed capacity [gFP MW−1]933927986963946964972984954962959
uranium content in spent fuel per unit of installed capacity [kgU MW−1]26.326.227.426.826.325.225.425.825.225.326.0

Over the period examined, the natural uranium SSI varied from 33 % (in 2007) to 58 % (in 2005). Low SSI values are evident between 2007 and 2010 (Figure 3), which indicates a heavy reliance on imports in these years. However, in 2011 there was a sharp increase in the SSI as domestic uranium prospecting and exploration intensified. This was because of an increase in investment from $ 89 m in 2010 to $ 131 m in 2012.[7] This is also reflected in the total drilling footage, which increased from 624 km in 2010 to 923 km in 2012. Similarly, the total number of holes drilled reached 4700 in 2012, up from 2219 in 2011 and 1816 in 2010.[7] As a result, the extraction of uranium resources has increased dramatically, particularly in northern China, in addition to the old mining areas in the south.[7]

More details on the trends over the period for the various parameters, calculated using the equations in the section on the estimation of uranium flows in China for PWRs and PHWRs, are provided in Tables 5 and 6, respectively, and a summary of the average values for the whole period is given in Table 7. The consumption of natural uranium per unit of installed capacity for PHWRs was much lower than that for PWRs, with average values of 0.132 and 0.194 tU MW−1, respectively (Tables 5 and 6). This is because PHWRs do not use enriched fuel that requires a large amount of natural uranium to concentrate 235U from 0.7 to 3–4 %. This, in turn, means that a significant amount of natural uranium is ultimately lost during the enrichment process as DU. Therefore, to minimize the consumption of natural uranium, PHWRs are a better alternative to PWRs. However, they generate much more spent fuel per unit of installed capacity than PWRs because they have a low burn-up rate: ≈7 GWd t−1 compared to an average of 35 GWd −1 for PWRs (Table 3). Consequently, PHWRs extract less energy from the fuel than PWRs, which results in a greater waste production per unit of electricity generated. Therefore, if the aim is to reduce the amount of spent fuel, PWRs are a better option. On the other hand, PHWRs provide fuel cycle flexibility as they can use different sources of fuel, which includes spent fuel from other reactors and DU from the enrichment process. However, the ultimate choice of an appropriate reactor type will depend on many other factors, including costs, the discussion of which is beyond the scope of this study. Instead, we now look at the potential flows of uranium in China in the near future.

Table 6. Fuel consumption and other parameters for PHWRs in China over the period 2003–2012.
installed capacity [MW]1456145614561456145614561456145614561456
power generation [GWh]788410 53410 12411 45811 54111 23811 72311 41211 50111 627
natural uranium consumption [tU yr−1]189.9180.1173.6196.5197.9192.4201.0195.7197.2198.8
natural uranium consumption per unit of installed capacity [kgU MW−1]131124119135136132138134136137132
spent fuel generation [t yr−1]183.3173.9167.6189.7191.0185.7194.0188.9190.4191.9
spent fuel generation per unit of installed capacity [kg MW−1]126119115130131128133130131132128
Table 7. Parameter summary for PWRs and PHWRs (average values over the period 2003–2012).
  1. [a] Data not available as PHWR spent fuel is not considered for reprocessing.

natural uranium consumption per unit installed capacity [kgU MW−1]194132
separative work unit per unit installed capacity [kg SW MW−1]131
uranium fuel consumption [kgU MW−1]27.6
spent fuel generation per unit installed capacity [kg MW−1]27.2128
uranium content in spent fuel per unit installed capacity [kgU MW−1]26.0NA[a]
plutonium content in spent fuel per unit installed capacity [gPu MW−1]272NA[a]

Uranium flows in the near future

The near-term future needs for natural uranium and the generation of spent fuel up to 2020 shown in Figure 4 were estimated using the projected installed capacity of electricity plants in China (Table 1) and the parameters listed in Table 7. As can be seen, the installed capacity is expected to reach 58 GW in 2020 (Table 1). Consequently, the consumption of natural uranium for power generation and the amount of spent fuel in 2020 will be 11 134 and 1725 tU, respectively (Figure 4), excluding the fuel used for the first loading of new reactors. This is 4.8 times higher than the amount of fuel used in 2012 and 3.5 times greater than the amount of spent fuel produced that year. In addition, the natural uranium demand for the first loading of new reactors will vary from a peak value of 3819 tU in 2017 to a low of 440 tU in 2018, according to the capacity of new reactors expected to come online.[4] The SSI is likely to decrease to 10–15 % in the 2016–2020 period if the domestic natural uranium production levels off at 1500 tU yr−1. Thus, the total natural uranium requirements from 2012 to 2020 would increase by a factor of 5.5 (Figure 4). For context, the reasonably assured and inferred conventional natural uranium resources in China were 265 500 tU in 2013 (Table S4). If the average annual increase in demand of 16.3 % from 2016–2020 were to continue in the future, this resource would be expended by 2027, 11 years from now (based on a starting year of 2013, contemporaneous with the resource estimate; for details, see Table S5). Given that new reactors are expected to operate for 50 years and possibly beyond, this suggests a serious shortage of uranium resources in China if business continues as usual.

Figure 4.

Estimated natural uranium demand and spent fuel generation based on the installed nuclear plant capacity in the period 2012–2020.

However, we should bear in mind that these figures are based on estimates using the current fleet of operating reactors in China, which introduces some uncertainty. The average burn-up of current plants is ≈7 GWd t−1 for PHWRs and 35 GWd t−1 for PWRs (Table 3). New nuclear build elsewhere in the world aims for higher burn-ups, often exceeding 50 GWd t−1 (see, for instance, Ref. [30]), which reduces the natural uranium use per unit of electricity generated and changes the mass and composition of spent fuel per unit of electricity generated. However, there is a trade-off that results from the reduction in uranium demand: a higher burn-up is typically achieved by increasing the enrichment of fuel to 4–5 % (as opposed to 3–4 % discussed above). As identified in the section on uranium flows from 2003–2012, enrichment is the main cause of natural uranium consumption in the PWR fuel cycle; therefore, to some extent, a higher enrichment counteracts the savings gained by increasing the burn-up. The balance between these variables and the effect they have on the values shown in Figure 4 will depend on the operational characteristics and management of nuclear new build and are difficult to predict until those new reactors are operational.

China is expected to have 58 GW of nuclear power capacity by 2020, which will place it in the third place in the world after the USA and France. In addition, by that time the country should also have an additional 30 GW or more of new nuclear generating capacity under construction.[4] With this expansion of capacity, the demand for natural uranium and the discharge of spent fuel will become even larger. Thus, it is essential for China to use natural uranium efficiently and to expand fuel-reprocessing technology extensively to improve the efficiency of the nuclear fuel cycle and become less reliant on imports, in line with some other countries. For example, France has the greatest capacity for the treatment of spent fuel from LWRs (1700 t yr−1 out of the 5400 t yr−1 of global reprocessing capacity).[31] The UK too has an extensive reprocessing capacity, partly for legacy fuels but also for LWR fuel, while Russia, Japan, and India also reprocess to some extent. While France uses the plutonium extracted from spent fuel to manufacture MOX for use in its own reactors, the UK primarily reprocesses LWR spent fuel for foreign customers as no MOX is used in domestic reactors. However, even in the case of France, despite extensive reprocessing activity, only ≈20 % of nuclear electricity is ultimately generated from recycled spent fuel.[32] Globally, the contribution of reprocessed fuels to the nuclear power industry remains quite low. If the aforementioned Chinese industrial reprocessing plant that treats 800–1000 t yr−1 of spent fuel is in operation by 2025, it will produce 115–140 t yr−1 of MOX (based on 1 % of Pu in the PWR spent fuel and 7 % in the MOX). As the MOX and uranium fuels are similar in terms of the energy they provide, this would help to reduce the natural uranium demand by around 840–1030 t yr−1 (estimated by Equation (3) by assuming an average value of 3.5 % for XEn and 0.25 % for XDU). However, given that the total nuclear power capacity is expected to reach 58 GW in 2020, around 487 t yr−1 of MOX will be required, which necessitates the reprocessing of 3400 t yr−1 of spent fuel (for the estimates, see Section S2 in the Supporting Information). Therefore, the planned reprocessing capacity will not be sufficient and would need to be further increased to reduce the use of natural uranium and the amount of nuclear waste. However, any future expansion of reprocessing will depend on the economic situation at the time, particularly the costs of the production of MOX in comparison to the price of natural uranium.


We have presented a comprehensive uranium flow analysis for China's nuclear power industry over the ten-year period from 2003 to 2012, with a projection of uranium needs and spent fuel generation in the near-term up to 2020.

The results show that in 2012, China's operating nuclear power plants consumed 2341 tU of natural uranium (excluding the amount used for the first loading of new reactors); this is double the amount consumed in 2003. The cumulative consumption over the period was 19 126 tU, which includes 3236 tU used for the loading of new reactors. Much of that was lost during the enrichment process as DU, which amounted to 11 890 tU. In total, 3911 tU of nuclear fuel was consumed, which generated 3811 t of spent fuel.

The natural uranium self-sufficiency varied over the period from 33 to 58 %. It was particularly low between 2007 and 2010, before increasing sharply in 2011 as domestic uranium prospecting and exploration intensified. The efficiencies of uranium fuel use and spent fuel generation increased in 2004, 2007, and 2011 because more efficient units came online.

China's installed capacity of nuclear power will increase rapidly, particularly in the short term, with approximately 10 GW of new build added annually, expected to reach a total installed capacity of 58 GW by 2020. Consequently, the annual consumption of natural uranium and the generation of spent fuel is expected to amount to 14 426 and 1725 t, respectively. Based on the rate of increase in demand, the domestic natural uranium resources could be expended by 2027. Thus, the future supply of natural uranium will have to come either from the reprocessing of fuel or through imports. To improve self-sufficiency and minimize natural uranium consumption and imports to 2020 and beyond, reprocessing deployment rates should be increased by a factor of 3–4.

China currently has the biggest program of nuclear construction in the world and, in the coming decades, it is likely to overtake France and eventually the USA to become the world's leading producer of electricity from nuclear sources. As discussed above, a recent surge in domestic extraction has improved China's self-sufficiency with respect to uranium. However, it is not clear to what extent domestic production, and indeed international production, can grow to accommodate the anticipated expansion in power capacity. Therefore, at this stage of the development of its nuclear industry, China is in a unique position to explore various fuel cycle options, including reprocessing, different enrichment schemes, and the exploitation of the ability of non-PWRs to accommodate alternative fuel sources. This provides many possibilities to minimize natural uranium consumption and waste generation before the nuclear fleet becomes locked into specific fuel cycles because of existing capital investment.

While in this paper we established the current and near-term state of several fuel flow indices, a forthcoming study will build on this work to explore some of the future possibilities to optimize the utilization of uranium in China.



fraction of component i in spent fuel

Bd [MWd tU−1]

average discharge burn-up


load factor

EFuel [kWh gU−1]

efficiency of nuclear fuel use

ESF [kWh g−1]

reactor efficiency

FP [tFP]

fission products in spent fuel

LConv (0.005)

rate of uranium loss in the conversion process

LFabr (0.005 each)

loss rates in the production of UO2 and the final reactor fuel


loss rate of uranium fuel through the burn-up

MConv [tU]

input of UF6 in the enrichment stage

MDU [tU]

amount of depleted uranium

MEn [tU]

amount of enriched uranium (3–4 % 235U)

MFuel [tU]

uranium fuel consumption

Mi [tU, tPu, or tFP]

amount of component i in spent fuel

MLoad [tU]

amount of natural uranium in the first loading of new reactors

MNat [tU]

amount of natural uranium

MNat (Dom) [tU]

amount of natural uranium produced domestically over a certain period

MSF [t]

amount of spent fuel


installed thermal power

P [MWh yr−1]

annual power generation


uranium recovery rate during the natural uranium conversion, production of UO2, and fuel fabrication, equal to 0.995 in each step

S [tSW]

separative work unit (SWU) requirement

SSI [%]

self-sufficiency index

XDU [0.2–0.3 %]

proportion of 235U in depleted uranium

XEn [3–4 %]

proportion of 235U in enriched uranium

XNat [0.7 %]

proportion of 235U in natural uranium


This research was supported by the UK Research Councils (EPSRC and ESRC, Gr. no. EP/F001444/1), the China Scholarship Council (CSC) and the National Natural Science Foundation of China (71373003).