The molten salt reactor (MSR) (Fig. 13.12) is the most innovative of the potential systems identified by GIF, departing significantly from current technologies. The MSR was first developed in 1954 by the US military. Further research was undertaken in the US in the 1960s, with two demonstration reactors built at the Oak Ridge National Laboratory (Abram and Ion, 2008; Renault et al., 2009). Instead of solid fuel elements, the MSR system uses a circulating salt mixture that contains the fissile material, typically a liquid mixture of fluorides of sodium, zirconium and uranium, which acts as both fuel and coolant. It circulates continuously through a graphite core, and then through a heat exchanger, where it transfers heat to a secondary salt circuit. A proportion of the salt is then diverted through a processing plant where fission products are removed and new fissile material is introduced. This continual processing of the fuel allows operation without refuelling.

Control rods

Coolant

©

Emergency dump tanks

73.72 Molten salt reactor (MSR).

The design of the MSR means that it is a closed cycle system, which makes efficient use of U, Th, Pu and MAs. This design also has a unique safety feature, since the fuel can be easily drained from the reactor in the event of an accident (Abram and Ion, 2008). Compared with solid-fuel reactors, MSR systems do not need fuel elements, have lower fissile inventories, lower radiation damage that can limit fuel burn-up, efficient actinide burning, a mechanism for continuous FP removal, the possibility of adding makeup fuel as needed, which precludes the need for providing excess reactivity and extends fuel resources, and a homogeneous isotopic composition of fuel in the reactor (Schenkel et al., 2009).

However, MSR systems present a number of challenges. The processing of the highly radioactive salt mixture requires robust processing equipment and materials. The combination of a corrosive and radioactive salt, with an isotopic composition, which changes over time, and a high neutron fluence also places extreme requirements on the primary circuit components. The graphite core also receives a high radiation dose, which means almost certain replacement during the life of the reactor. These challenges mean that the MSR system is the least developed system with less chance of achieving commercial viability by 2030 than other designs (Abram and Ion, 2008). However, given the potential of the system, the MSR provisional system steering committee (PSSC) of GIF decided in 2008 to modify the Generation IV Roadmap agreed in 2002 to include research on fuel and coolant salts.

There is a range of developments that may help in the development of MSR systems, including Brayton power cycles (rather than steam cycles), which eliminate many of the historical challenges in building MSRs (Schenkel et al., 2009). Two main design concepts have subsequently been developed (Renault et al., 2009):

1 The fast-spectrum MSR (MSFR) as a long-term alternative to solid-fuel fast neutron reactors, characterized by large negative temperature and void reactivity coefficients, a unique safety characteristic not found in solid-fuel fast reactors.

2 The fluoride-cooled high-temperature reactor (FHR), a high-temperature reactor, which is more compact than the VHTR and has the potential for passive safety from small to very high unit power (>2400 MWft).

Developments in MSFR and FHR are discussed in more detail in the following sections.

Research has also been undertaken into the use of liquid salt technology in other reactor system (both nuclear and non-nuclear), including other Generation IV systems (Schenkel et al., 2009). Liquid salts could be used, for instance, as primary coolants in an FHR, as an alternative to secondary sodium in sodium fast reactors (SFRs), and to intermediate helium in VHTRs. Investigations into high — temperature salts as coolants may lead to other nuclear and non-nuclear applications. Possible examples include heat transfer for nuclear hydrogen production, concentrated solar electricity generation, oil refineries and shale oil processing facilities. Liquid salts have two key advantages (Renault et al., 2009):

1 their higher volumetric heat capacity allows for smaller equipment size

2 the absence of chemically exothermic reactions between them and the coolants used in the intermediate loop and power cycle coolants

There are a number of research priorities in improving understanding of liquid salt chemistry, including (Schenkel et al., 2009):

• the physico-chemical behaviour of coolant and fuel salts, including fission products and tritium

• the compatibility of salts with structural materials for fuel and coolant circuits, as well as fuel processing material development

• the on-site fuel processing for MSFR; the maintenance, instrumentation and control of liquid salt chemistry (redox, purification, homogeneity)

• safety aspects, including interaction of liquid salts with sodium, water, air, etc.

Forecasts of the future spent fuel quantities are influenced by two or three major factors. The first is the projected growth of nuclear energy use. The International Atomic Energy Agency (IAEA) periodically updates its projections of global growth in nuclear energy use.1 As in previous years, the 2010 Nuclear Technology Review presented low and high estimates of nuclear capacity in 2030; these were 511 and 807 GW(e) respectively. Both figures were around 8% higher than the estimates presented in the previous year and they continued a generally increasing trend over the decade. This well illustrates the inherent uncertainties in using these figures to derive estimates of future arisings of spent fuel. The other factor is the fate of spent fuel discharged from power reactors. No facilities for disposal of commercial spent fuel were available in 2010 so that spent fuel could either be stored or reprocessed. The total amount of nuclear fuel discharged by that year was approximately 320 000 tonnes of heavy metal (t HM). Of this amount approximately 95 000 t HM had already been reprocessed and about 225 000 t HM was stored either in at-reactor storage pools or in away-from — reactor storage facilities (dry or wet storage technology). So in addition to the installed nuclear capacity, the actual quantities of stored fuel will also depend on whether fuel is reprocessed or not. Reprocessing capacities in 2010 were 5000 t HM per year with another 1000 t HM capacity expected to start operation soon at Rokkasho in Japan. The last factor that may change the accumulated quantities of spent fuel is the introduction of advanced fuel cycles such as MOX or higher burn up fuel. Figure 15.1 shows a prediction of spent fuel arisings until 2020 and shows the distribution between storage and reprocessing. The figure assumes that total nuclear generation in 2020 will be about 420 GW(e), which is 13% higher than 2009.

Figure 15.2 shows the cumulative fuel discharges up to 2010 by country. It omits reprocessed fuel.

Uranium purification

The purification step enables the ‘cleaning’ of uranium and plutonium from the fission products and short-term noxious actinides (Fig. 16.9).

The purification of uranium is performed using liquid-liquid separation in sets of extractors-settlers. Zirconium, thorium, ruthenium (fission products, /emitters) and plutonium and neptunium traces still there are then removed and transferred to the vitrification workshop. Uranium is concentrated (using evaporation) and recovered as uranyl nitrate. Some the uranyl nitrate is used to make the uranous nitrate during the extraction step previously described. The remaining uranyl nitrate can be re-used in a new cycle.

Stripping (de-extraction) returns uranium to the aqueous phase. This is completed by washing the solvent rich with uranium with a high flow rate of slightly acidic water at around 50 °C. These conditions are unfavourable for the stability of the uranium nitrate complexes.

After concentration, uranium is submitted to an additional purification cycle to remove neptunium (Np), the main pollutant of uranium.

Conversion to Pu IV

The processing of the plutonium solutions from the extraction step follows two routes.

The first route is a wet process, which is a chemical treatment. This treatment targets the final production of plutonium oxide.

The second route is a dry process and is a mechanical treatment. This enables the conditioning of the plutonium oxide produced and shipping to MOX production.

Plutonium from the partition in solution it is at valence III. It is oxidized to level IV by spraying the liquid (‘sparging’) into a nitrous vapour.

Hydrazine nitrate is reduced while plutonium is oxidized:

2 Pu3+ + NO3- + 3H+ ^ 2Pu4+ + HNO2 + H2O

Reloaded with acid, the solution can be retreated with additional extraction/ scrubbing cycles.

Remarkable progress in the last two decades has been made in the field of geological disposal. Some countries have reached important milestones and geological disposal (of spent fuel) is expected to start in 2020 in Finland and in 2022 in Sweden and in fact the licensing of the geological repositories in both countries is now entering into its final phases. In France disposal of ILW and vitrified HLW is expected to start around 2025, according to the roadmap defined by a Parliament Act in 2006.

The impact of P&T technology on the management of nuclear waste has been discussed since the 1970s. Up to about the year 2000, the emphasis was on reductions in the ‘potential radiological toxicity’ or ‘radiotoxic inventory’, which is defined by the sum of the activities of individual radionuclides in the waste when normalized by their dose coefficients for ingestion.1 One of the important studies in this period was the first Status and Assessment Report on P&T issued by the OECD/NEA in 1999.29 The reduction of radiotoxicity was considered to be important for the long-term safety of deep repositories, especially for disturbed evolution scenarios, such as human intrusion. In the decade following the year 2000, more emphasis was put on the realistic consideration of issues pertaining to the fuel cycle and to the implications of P&T for repository design.20-41 Results have been summarized and discussed by an OECD-NEA Task Force.42

A comprehensive description of all aspects of geological disposal is:

Geological Repository Systems for Safe Disposal of Spent Nuclear Fuels and Radioactive Waste, eds. Joonhong Ahn and Michael J. Apted, Woodhead Publishing 2010.

Somewhat older but still useful is:

The Scientific and Regulatory Basis for the Geological Disposal of Radioactive Waste, ed. David Savage, John Wiley 1995.

Without exception national waste management organisations are committed to openness and transparency. Consequently, it is normal for them to provide a vast amount of information, including reports at every level of detail, on their websites. Similarly, the IAEA provides free downloads of its technical publications on its website. Some useful addresses are provided in Table 18.2.

6.1.1 Central Asia’s uranium mines

Kazakhstan

Kazakhstan has been an important source of uranium for more than fifty years. Uranium exploration started in 1948 and economic mineralisation was found in several parts of the country. This supported various mines in hard rock deposits. Some 50 uranium deposits are known, in six uranium provinces. In the early 1970s, successful tests on in situ leaching (ISL) led to further exploration being focused on two sedimentary basins with ISL potential. Up to 2000, twice as much uranium was mined from hard rock deposits as sedimentary ISL sands, but almost all production is now from ISL mines, some relatively small. Over 2001-2011 production rose from 2000 level to 19 450 tU/yr, making Kazakhstan the world’s leading uranium producer (36% of total in 2010), and further mine development is under way with a view to increasing production. All uranium is exported.

Kazatomprom is the national atomic company set up in 1997 and owned by the government. It controls all uranium exploration and mining as well as other nuclear-related activities. It aims to add value to the fuel chain and it is developing its fuel fabrication facilities so that fuel assemblies, rather than just fuel, could account for most sales by 2015. Kazatomprom has forged many international agreements on all aspects of nuclear power, and many of the mining operations are run as joint ventures with Russian, Chinese, Canadian and French companies.

All except one of the operating and planned ISL mines are in the central south of the country. Mines in the Stepnoye area have been operating since 1978, those in the Tsentralnoye area since 1982 — both in the Chu-Sarysu basin/ province, which has more than half the country’s known resources. There are 14 mines here. Mines in the Western (No. 6) area of the Syrdarya basin/province have operated since 1985, and today it has seven mines. One further ISL mine is in the Northern province.

Russia

AtomRedMetZoloto (ARMZ) is the state corporation which took over all uranium exploration and mining assets in 2007, as a subsidiary of Atomenergoprom, the state-owned Russian atomic energy company. It inherited 19 projects with a total uranium resource of about 400 000 tonnes.

Uranium production is increasing. In 2010 Russia produced some 3500 tonnes of uranium, mostly from several large underground mines operated by Priargunsky in the Streltsovskiy district of the Transbaikal or Chita region of south-east Siberia near the Chinese and Mongolian borders. These deposits were discovered in 1967 and have been the major source of production since.

A lesser amount of production is from new operations at Khiagda in Buryatiya about 500 km north-west of Priargunsky’s operations, and Dalur in the Kurgan region between Chelyabinsk and Omsk, just east of the Urals. Both are low-cost in situ leach (ISL) operations.

Most of the future production is set to come from the massive Elkon project with several mines in the Sakha Republic (Yakutia) some 1200 km north-north-east of the Chita region. There is huge investment to bring these into production, which could ramp up from 2013 levels to 3000 tU in 2015, and 5000 tU/yr by 2024.

Uzbekistan

During the Soviet era, Uzbekistan provided much of the uranium for the Soviet military-industrial complex. Today the state-owned Navoi company operates several uranium mines, producing about 2400 tU/yr.

No major technical hurdles are foreseen in fabricating thorium fuels due to a quite extensive experience base extending back to the 1960s. More recently, in the context of EC programmes, Pu/Th-pellets were fabricated and irradiated in a BWR and no specific problems were encountered; indeed, thorium oxide fuels had superior irradiation behaviour compared to uranium oxide fuels (see for example, reference 3 pages 28 to 42). Nevertheless, additional fabrication and especially irradiation testing will be needed if burn-ups are to match today’s UOX and MOX experience, i. e. the 50-60 GWd/tHM range, and even beyond 70 GWd/ tHM for some thorium-fuel options considered.

More than 40 years of experience have also been accumulated on HTR fuel fabrication. These reactors are considered to be among the best candidates to accommodate thorium fuels, because of their relatively high conversion ratios (a result of good neutron economy). In this regard, the fuel fabrication processes developed for the Fort St Vrain and THTR power reactors could be the starting basis for defining a new manufacturing plant for other reactor types. As a matter of fact, these two reactors were industrial prototypes (for the 300 MWe class, see

Table 8.1) and, thus, their associated fuel fabrication facilities used semi-industrial processes.

On the other hand, it must be underlined that India has recently manufactured thorium fuel on a scale that is beyond the R&D developmental phase (cf. Section 8.2.3): 7 tonnes of pellets have been manufactured for their PHWR 220 Units at their Nuclear Fuel Complex (NFC), Hyerband; and at BARC, 5 tonnes have been manufactured for their liquid metal cooled FNR programme.

Although development of a thorium-based fuel cycle would require significant additional work to bring it to an industrial scale, there are no major technical hurdles to the manufacture of thorium-based fuels. The experience gained from LEU fuel would provide a base line for the development of this fuel fabrication process.

In a PWR the steam generator is the interface between the reactor coolant, which will be active, and the steam, which drives the main turbines. Its duty is to transfer heat from the reactor coolant water to feedwater from the turbine condensate system and to convert it to nearly dry steam. Each loop of the reactor coolant system contains a vertically mounted U-tube steam generator, as shown in Fig. 10.7. The steam generator consists of three sections, a hemispherical bottom head carrying

the primary coolant inlet and outlet nozzles, an evaporator section enclosing the U-tube bundle, and an upper section enclosing the moisture separators.

The reactor coolant flows through the inverted U-tubes, entering and leaving through the nozzles located in the hemispherical bottom head. The head is divided into inlet and outlet chambers by a vertical partition plate. The Inconel heat transfer tubes are hydraulically expanded into the forged tube sheet and the ends are seal welded to the tube-sheet cladding. The tubes are supported at intervals by stainless steel horizontal support plates, which have clearance holes to permit flow of the steam/water mixture.

Feedwater flows directly from the feedwater distributor ring down through the annular space between the shell and the cylinder surrounding the tube bundle. It then passes in to the bottom of the tube bundle where boiling commences. The resultant water/steam mixture flows upward through the tube bundle and into the upper section. A set of centrifugal moisture separators located above the tube bundle removes most of the entrained water, which passes to the annular space to mix with the incoming feedwater. Steam driers above the moisture separators increase the steam quality to a minimum of 99.75% (i. e. the steam contains not more than 0.25% water). The dried steam passes through a nozzle in the upper dome to one of the four main steam pipes.

Reactor control was by use of boron-steel or steel control rods. Boron steel was typically used for shut-down or bulk control, whilst mild steel ‘grey rods’ were used for trimming or power regulation. Typical rod design was as a series of about five articulated sections to permit insertion even in the event of disruption or movement of the core graphite. As is the case with many Magnox features, each station used a unique design of control rod operation and drive mechanism.

Further protection was provided to enable emergency shut-down in the form of boron ball shut-down systems. These comprised hoppers of boron-steel balls located above a thimble in an interstitial channel. Opening of the hopper dropped the balls into the core, inserting a large amount of negative reactivity. Operation of the boron ball shut-down mechanism was not terminal to reactor operation, and the devices were tested on a routine basis. The boron ball devices had a high degree of resilience against core disruption.

Boron dust injection mechanisms were retro-fitted as a final means of holding down the reactivity post-trip to prevent re-criticality. The mechanism for injecting boron carbide and boron trioxide powders in an air flow would have been manually connected to the reactor (preventing inadvertent automatic deployment). Use of the boron dust system would have irreversibly shut down the reactor.

Thermal expansion occurs instantaneously as the fuel temperatures increase. Since the fuel pellets/bars are at significantly higher temperatures than the cladding, differential thermal expansion between fuel pellets/bars and cladding leads to dimensional changes, which tend to close any pellet-cladding, or bar­cladding, gap. In the case of ceramic fuel pellets, the hotter central regions of the pellets expand more than the cooler outer regions, giving the pellets a distinctive ‘wheatsheaf’ or ‘hourglass’ shape. The differential thermal expansion within the fuel pellets also imposes shear stresses, which cause cracking of the pellets. The

resulting pellet fragments can relocate into any pellet-cladding gap. An exaggerated schematic of an idealised cracked and wheatsheafed pellet is reproduced in Fig. 14.1 (Gittus, 1972).

Both pellet-cladding, or bar-cladding, gap closure and fuel fragment relocation decrease the maximum fuel temperature at any given power.