End plugs or end caps close the cladding tube at both ends of the fuel rod and are made of a zirconium alloy. They engage with the end plates of the assembly and are exposed to less severe operation conditions since they are not subjected to mechanical stress by expanding pellets and run at lower temperature than the cladding (no heat flux). End plugs are forged and machined from blanks, which are cut from plate or barstock material. Depending on the rod assembly process, the upper end plug may have a hole that is welded tight after the rod is filled with helium.

9.4.1 Plenum spring

The plenum spring secures the fuel pellet stack against axial movement during transport and handling. It is made of one piece of stainless steel (BWR) or Inconel (PWR) wire. During initial operation, it counteracts the formation of axial gaps in the pellet stack by pressing the pellets downwards. The hold-down force is of the order of 4 to 5 times the weight of the fuel stack.

The ability to use natural uranium fuel was one of the principal requirements in the design of the CANDU reactor. This necessitates a design in which high neutron economy is of paramount importance, and is achieved by:

• minimizing parasitic absorption through design and through the use of low neutron absorbing structural materials in the fuel and core

• using heavy water (D2O) as both coolant and moderator, which avoids parasitic absorption in water and provides excellent neutron thermalization (e. g., slowing down of neutrons to energies where they are more likely to be absorbed by fissile isotopes)

• refuelling the reactor on power, which results in very low excess reactivity in the core, obviating the need for burnable poisons and minimizing neutron absorption in control materials. Low core excess reactivity also results in safety benefits

• employing a pressure tube design, which helps reduce resonance absorption in U-238 while achieving a very thermal neutron spectrum. (Grouping the fuel elements into a cluster (fuel bundle) surrounded by a large amount of heavy water moderator increases the probability that a high-energy neutron produced through fission will escape from the fuel into the moderator before being captured by the intermediate and high-energy neutron resonances in U-238.) A pressure tube reactor also avoids the difficulty of manufacturing a large pressure vessel

A schematic of a CANDU power plant is shown in Fig. 11.1. The reactor consists of a horizontal, stainless steel cylindrical tank called a calandria, which is filled with heavy water moderator. Inside the calandria are several hundred fuel channels running the length of the tank, arranged in a square lattice with a pitch of 28.6 cm. The fuel channels consist of a zirconium-alloy pressure tube inside a thinner zirconium-alloy calandria tube, separated by an insulating gas gap. Twelve fuel bundles, cooled by the D2O coolant at high temperature (around 300°C) and pressure (around 10 MPa), sit inside each pressure tube (see Fig. 11.2). The fuel channel separates, and insulates, the unpressurized, cool moderator on the outside of the channel from the pressurized, hot coolant within. The CANDU reactor design is modular, in that the number of fuel channel assemblies can be varied to give the desired power output. The CANDU 6 reactor has 380 fuel channels and, therefore, 4560 fuel bundles.

The coolant is pumped through a large inlet header into small inlet feeders, which are connected to one end of the fuel channels. The coolant removes fission heat from the fuel bundles as it moves through the channels, and flows through outlet feeders to a large outlet header, which is connected to a steam generator where the heat from the D2O coolant is transferred to ordinary water on the other side of the steam generator tubes, which turns to steam. The coolant is then pumped into an inlet header at the other end of the reactor. Figure 11.3 illustrates

Heavy water coolant (in heat transport system)

©

Cooling water from lake/ocean/river

Fuel ‘Heavy water Fuelling (uranium) moderator machine (in calandria)

77.7 Schematic of CANDU power plant (figure is copyright Atomic Energy of Canada Limited and is used with permission).

11.2 CANDU reactor fuel channel arrangement (figure is copyright Atomic Energy of Canada Limited and is used with permission).

11.3 Face view of reactor showing feeders (figure is copyright Atomic Energy of Canada Limited and is used with permission).

the face view of the reactor, showing the feeders. In a CANDU 6 reactor, the coolant flows in two separate figure-of-eight loops, each serving half of the fuel channels, with two heat transport pumps, two inlet and outlet headers and two steam generators at each end of the core. Coolant flow is bi-directional, meaning that coolant flow in adjacent channels is in opposite directions. The reactor core, primary heat transport pumps, steam generators and associated equipment are located in a containment building. The balance-of-plant, or secondary side, is outside the containment building, with the principal components being the steam lines, steam turbines, electrical generator, condenser and feed water lines to the steam generators. A summary of the parameters of the principal types of heavy water reactor is given in the Appendix in IAEA (2002).

A key concept in the Generation IV Initiative is a mix of nuclear reactor designs, where the strengths of some reactors counterbalance the weaknesses of others, in order to achieve fully sustainable nuclear energy production. That is

particularly true as far as nuclear waste transmutation is concerned. A few different integrated cycles have already been proposed. Generally, they are rather complex cycles, starting from the current Generation III or advanced Generation III+ (e. g. EPR, AP-1000, ABWR) LWR range of reactors. The LWR SNF could, for example, provide feed fuel for emerging Generation IV reactors. Each cycle has its advantages and drawbacks, and an ideal solution has not yet been found. In addition, due to the extreme complexity of MA behaviour in terms of core kinetics, some integrated cycles also envisage a dedicated subsidiary process using ADS (NEA, 2006b). Some additional examples are shown in Fig. 13.4 . Three kinds of possible fuel cycle can be highlighted (Bomboni, 2009):

1 Cycles based on ‘current industrial technology and extensions’: only LWRs and, if necessary, CANDUs are involved and only one recycle of HM is envisaged.

2 ‘Partially closed fuel cycles’: these cycles are fully closed only for Pu; in some schemes a single recycle of some MAs is envisaged in LWRs or FRs.

3 ‘Fully closed fuel cycles’: all the advanced reactor concepts, ADS included, could be involved; only HM losses and FPs go to the geological repositories; pyrochemical reprocessing is envisaged.

For the complete transmutation of HMs all actinides are recycled continuously in a closed fuel cycle until they fission (Bomboni, 2009). A closed fuel cycle cannot

be achieved without multiple recycling of all HMs. An example of an advanced fuel cycle, based on Generation IV reactors, which may maximize the exploitation of natural resources, minimize the final mass and radiotoxicity of the waste and be proliferation resistant, is the symbiotic LWR-VHTR-GFR cycle (Bomboni, 2009). Although further analysis is still required, this example shows the potential of a kind of symbiotic cycle involving two of the most promising Generation IV reactor concepts, VHTR and GFR. It does not aim at being ‘the’ solution, but it should be considered as an interesting, reasonably feasible possibility that offers some useful advantages.

System thermal-hydraulics codes evaluate the thermal-hydraulics of the entire primary circuit (unlike core thermal-hydraulics codes, which only evaluate the thermal-hydraulics in the core). The thermal-hydraulics of the secondary circuit may also be evaluated, since this can affect the coolant enthalpy. The core is generally modelled as in core thermal-hydraulics codes, but with a lesser degree of discretisation. The remainder of the primary circuit (and possibly also the secondary circuit) are divided into various components, which are connected by sections of pipework.

The fluid mass, momentum and energy conservation equations are solved for each component and pipework section using similar techniques to those described for core thermal-hydraulics codes, but with sub-models for specific components such as coolant pumps, valves, etc., which allow their effects on the working fluid to be calculated without explicit modelling of the components themselves. In order to simulate the effects of automated, or reactor operator initiated, control of components, a control system model is also required. (This can, for example, simulate automated opening of pressure relief valves when the fluid pressure exceeds a trip setpoint.) Finally, a transient neutronics module is often integrated into the code to model the coupling between the core neutronics and the core thermal-hydraulics during certain events (e. g. a steamline break in a PWR or power-flow oscillations in a BWR); alternatively, the system thermal-hydraulics code can be coupled to a whole core neutronics code.

Both steady-state and transient simulations are generally possible with system thermal-hydraulics codes. A typical application is the determination of the evolution of the fluid boundary conditions in the core for a specific fault, such as a loss of coolant accident (LOCA) in an LWR. The results can then be fed to a core thermal-hydraulics code for more detailed analysis of the thermal-hydraulics in the core, or directly to a transient fuel performance code for thermo-mechanical analysis of the fuel pins.

Reprocessing is a large-scale industrial enterprise that entails handling spent nuclear fuel in quantities of the order of 1000 tonnes/year. While it is highly desirable to use a continuous process, this is not simple because of a specific nuclear industry constraint: the danger of criticality within the dissolver given that it contains both fissile material and water (a neutron moderator). To avoid this there are three main restrictions:

• the handled mass must be limited; unfortunately this is too severe a constraint for an industrial operation involving plutonium

• the volumetric concentration of fissile material must be limited; this is impossible to achieve in the course of dissolution

• the capacity of the dissolver must be limited and must use a favourable geometry

An example of an appropriate solution is illustrated by the continuous rotary dissolver used by the AREVA reprocessing plants in La Hague. This device allows four actions to be performed simultaneously: (1) the loading of the pieces of fuel to be dissolved, (2) dissolution itself, (3) draining of the solution and (3) the unloading of the empty pieces of cladding (hulls). The dissolver includes a wheel with buckets rotating step by step in a flat tank containing the hot nitric acid.

Figure 16.4 illustrates the process: the hulls fall into the dissolver filled with hot concentrated nitric acid. The nuclear material is dissolved and separated from other components. The structural elements of the spent fuel (hulls, end-pieces) are removed for storage and compaction.

Gaseous effluents are treated, washed and filtered while the acid solution is clarified. Centrifugation (Fig. 16.5) separates the small particles (shearing fines) and insoluble fission products from the solution of uranium, plutonium and fission products in nitric acid.

After clarification, the solution contains:

• 200 g/l uranium

• 2.5 g/l plutonium

• 3.5 mol/l nitric acid

• between 6 and 7 g/l fission products

The fines are stored for later vitrification and the clarified solution is transferred to the extraction and concentration plants where the nuclear materials are separated.

How P&T is to be implemented very much depends on what one wishes to achieve and this will vary from one country to another according to national policy. Nevertheless, three broad objectives can be identified:

• sustainable development of nuclear energy and waste minimization

• reduction (elimination) of MA inventory

• reduction (elimination) of TRU inventory as unloaded from LWRs

I n all cases, the objectives lead to a significant reduction of the burden on a geological repository (see Section 17.5). As possible ways of achieving these three objectives, the following three generic scenarios have been proposed.24 All of them go beyond the strategy of the ‘once-through’ fuel cycle (i. e. final disposal of once irradiated fuel) and imply fuel reprocessing.

Work to identify suitable geological environments for deep disposal of radioactive waste began in the 1960s and, by the 1980s, many possibilities had been defined, namely bedded salt,30 salt domes,31 unsaturated rocks,32 basement under sedimentary cover (BUSC),3 3 hard rock in low relief terrain, deep sedimentary basins, seaward dipping and offshore sediments, low permeability formations and small islands in the sea.34 In all these cases the underlying aim is either to prevent water contact with the waste (as with salt and unsaturated sites) or else to minimise its effects. BUSC is an interesting (and not uncommon) example where the higher permeability of the sedimentary cover means that groundwater flow is directed away from the less permeable basement rock, making the latter a more suitable host for a repository. Further, the permeability of sedimentary rocks is often anisotropic, being higher parallel to the bedding plane. This effect will also tend to direct the flow into the upper layers.

Some orebodies lie in groundwater in porous unconsolidated material (such as gravel or sand) and may be accessed simply by dissolving the uranium and pumping it out — this is in situ leaching (ISL) mining (also known in North America as in situ recovery — ISR). It can be applied where the orebody’s aquifer is confined vertically and ideally horizontally. It is not licensed where potable water supplies may be threatened. Where appropriate it is certainly the mining method with least environmental impact.

I SL mining means that removal of the uranium minerals is accomplished without any major ground disturbance. Weakly acidified groundwater (or alkaline groundwater where the ground contains a lot of limestone such as in the USA) with a lot of oxygen in it is circulated through an enclosed underground aquifer, which holds the uranium ore in loose sands. The leaching solution dissolves the uranium before being pumped to a surface treatment plant where the uranium is recovered as a precipitate. Most US and Kazakh uranium production is by this method.

In Australian ISL mines the oxidant used is hydrogen peroxide and the complexing agent sulfuric acid to give a uranyl sulfate. Kazakh ISL mines generally do not employ an oxidant but use much higher acid concentrations in the circulating solutions. ISL mines in the USA use an alkali leach to give a uranyl carbonate due to the presence of significant quantities of acid-consuming minerals such as gypsum and limestone in the host aquifers. Any more than a few per cent carbonate minerals means that alkali leach must be used in preference to the more efficient acid leach.

In either the acid or alkali leaching method the fortified groundwater is pumped into the aquifer via a series of injection wells where it slowly migrates through the aquifer leaching the uranium bearing host sand on its way to strategically placed extraction wells where submersible pumps pump the liquid to the surface for processing.

Acid consumption in acid leach environments is variable depending on operating philosophy and geological conditions. In general, the acid consumption in Australian ISL mines is only a fraction of that used in a Kazakh mine (per kilogram of uranium produced). A general figure for Kazakh ISL production is up to 80 kg acid per kgU, though some mines are a bit lower. This is becoming a significant cost constraint there. Beverley in Australia is reported to be 3 kg/kgU.

For very small orebodies that are amenable to ISL mining, a central process plant may be distant from them so a satellite plant will be set up. This does no more than provide a facility to load the ion exchange (IX) resin/polymer so that it can be trucked to the central plant in a bulk trailer for stripping. Hence very small deposits can become viable, since apart from the wellfield, little capital expenditure is required at the mine and remote IX site.

8.1.2 Experience of thorium use in experimental and power reactors

During the pioneering years of nuclear energy, 1950-1970, with great enthusiasm and regardless of the costs, a large number of possible avenues for energy production with thorium were investigated, not only in the USA and USSR, but also in Europe and, to some extent, in Asia.[14] For example, it is remarkable that the thorium-based Elk River (1963) and Peach Bottom (1967) reactors were started only a few years after the ‘founding fathers’ of the two main reactor families of today, based on uranium fuel, PWR Shippingport (1957) and BWR Dresden (1960). It is also remarkable that breeder demonstration was performed at Shippingport in the late 1970s and early 1980s using a U-233/thorium cycle.3 The conversion ratio1 reached 1.0139. This was the only US demonstration programme using U-233 as the fissile seed material. Although this demonstration was successful from the standpoint that slightly more U-233 was bred than consumed, success was only achieved at the high cost of a sophisticated core design, and by sacrificing reactor performance.

From that time on, a significant amount of experience on thorium-based fuel in experimental and power reactors has accumulated. An exhaustive list of these reactors is provided Table 8.1 (however, this table does not include experimental reactors in which thorium fuels have also been tested, such as CIRUS in India, KUCA in Japan, MARIUS in France, etc.).

Within the framework of this chapter, it is not possible to provide details on all of these reactors and, instead, the reader is directed to references 3 to 6. Here we shall focus on high-temperature gas-cooled reactors (HTRs) since, as is seen in Table 8.1, thorium fuel was mainly developed for this type of reactor.

In the US, during the 1960s and 1970s, the use of a HEU (highly enriched uranium)-thorium fuel cycle was demonstrated at the Peach Bottom and subsequently, Fort Saint Vrain HTRs. Both reactors used prismatic block type fuel elements containing either fissile or fertile fuel. The fissile fuel consisted of HEU dicarbide, the fertile fuel was thorium dicarbide. Both fuels were in the form of

Country Name Type Power (MW) Startup date Fuel Note USA Indian Point 1 PWR 265 1962 Th02-U02 Power includes 104MWe from oil-fired superheater Elk River BWR 22 e 1964 Th02-U02 Power includes 5MWe from coal-fired superheater. Thorium loaded in the first core only Shippingport PWR 60 e 1957 Th02-U02 Used both U-235 and plutonium as the initial fissile material. Successfully demonstrated thermal breeding using the ‘seed/blanket’ concept (Th/U-233) Peach Bottom HTR 40e 1967 ThC2-UC2 Coated particle fuel in prismatic graphite blocs — Th/HEU Fort St Vrain HTR 330 1976 ThC2-UC2 Coated particle fuel in prismatic graphite blocs — Th/HEU MSRE MSR 1°,h 1965 ThF4-UF4 Operated with U-233 fuel since October 1968. No electricity production UK Dragon HTR 20th 1964 ThC2-UC2 Coated particle fuel. No electricity production. Many types of fuel irradiated Germany AVR HTR 15 e 1967 ThC2-UC2 Coated particle fuel in pebbles. Maximum burn-up achieved: 150 GWd/t — TH/HEU THTR HTR 300 e 1985 ThC2-UC2 Coated particle fuel in pebbles. Maximum burn-up achieved: 150 GWd/t — Th/HEU Lingen BWR 60 e 1968 Th/Pu Th/Pu was only loaded in some fuel test elements India Kakrapar (KAPS) 1-2 PHWR 200 e 1993/95 U02-Th02 Fuel: 19-element bundles. 500kg of Th loaded Kaiga 1-2 PHWR 200 e 2000/03 U02-Th03 Fuel: 19-element bundles. Th is used only for power flattening Rajasthan (RAPS) 3-4 PHWR 200 e 2000 U02-Th04 Fuel: 19-element bundles. Th is used only for power flattening KAMINI Neutron source 30 Kwe U-233 Experimental reactor used for neutron radiography

carbon-coated particles, the fissile particles being somewhat smaller than the fertile ones.

In the UK, the first HTR demonstration known as Dragon operated between 1966 and 1975. Various types of fuel elements including thorium with a 10:1 Th/U (HEU) ratio were irradiated.

In Germany, two pebble bed type HTRs were operated. The first one, AVR, was a prototype pebble bed reactor that mainly used a HEU/thorium cycle. The fuel consisted of billiard ball-sized fuel elements. A commercial version, the THTR — 300, a 300 MWe thorium/HEU fuelled HTR, started operation in 1985. It was permanently shut down in 1989 largely for political reasons although high operational costs and an operational incident in 1986 that resulted in the release of radioactive materials are often mentioned as the grounds for shutdown.

The main distinguishing feature of these types of plants is the use of water as both a moderator and a coolant. Although water is an effective moderator it also has a high neutron capture cross section. It can, however, be used as a coolant, but for electricity production it needs to be pressurised to produce the steam necessary to operate turbines. To overcome neutron absorption, slightly enriched fuel is used. LWRs are designed to be compact and slightly under-moderated.

The obvious difference between the two designs of LWR is the fact that in a PWR the pressure is sufficiently high that boiling in the core is suppressed. The pressure in a BWR is lower and so boiling occurs in the core itself. This has led to the use of a direct cycle in which the steam produced in the core is used to directly drive the turbine generator. Figures 10.1 and 10.2 show overall schematics of a ‘generic’ PWR and BWR. It should be noted that the core cooling water will become radioactive as a result of its passage through the core. Nitrogen-16 is produced by the activation of oxygen. This has a half-life of 7.13 seconds and so decays rapidly once the reactor is shutdown. In addition corrosion products may become activated and if there are fuel failures or tramp fuel is present then fission products will be present in the coolant water. The control of activity in the coolant will be discussed in Section 10.4.

The PWR consists of a primary circuit in which water at high pressure (typically —15.5 MPa) is circulated through the core to provide both cooling and moderation. The water enters the core at about 293 °C and leaves at —324 °C and then passes through a number of steam generators, which are water/water heat exchangers. Each of the subsystems consisting of a steam generator, its associated pipework and reactor coolant pump(s) is known as a reactor coolant loop. The secondary side of the steam generator is at a lower pressure (typically —6.9 MPa) so the water boils generating steam. The steam is saturated and is relatively wet so the upper section of the steam generator contains separators and driers. The steam then passes through the turbine and the condensate is returned to the steam generators by the main feed pumps.

As will be discussed below, although the early BWRs also had steam generators direct cycle plants were introduced in the early 1960s and modern plants were developed from these. The primary circuit is maintained at a lower pressure than is the case for a PWR (typically —7 MPa) and boiling occurs in the core region. The reactor pressure vessel is taller than that of a PWR because it contains steam separators and driers in the upper part of the vessel. The steam, which will contain some radioactive material, then passes through the turbine and the condensate is

Cooling tower

10.2 Schematic of a BWR (Source: USNRC).

returned to the vessel by the main feed pumps. In addition the current BWRs also have recirculation pumps. These take water from the pressure vessel downcomers, which are fed by both condensate and water draining from the separators. This is then pumped into the vessel lower plenum to increase the flow through the core.

Although the two reactor types have characteristics in common there are sufficient differences to make it more convenient to discuss them separately. However, before doing so it is worth saying a little about the overall operational performance requirements demanded of modern nuclear power plants.

In the early days of reactor development for electricity production the plants were designed on the assumption that they would operate as baseload stations (i. e. they would run constantly between maintenance and refuelling outages) with load following and frequency control being undertaken by conventional fossil fuelled plants. This is still the case in countries where the proportion of nuclear generation is still relatively low. However, in a number of European countries nuclear generation exceeds baseload. In addition as electricity markets have been deregulated and access to grid systems has been opened up, the grid system operators now require almost all plants wishing to connect to the grid to be able to contribute to its stabilisation by offering load following and frequency control.

In this respect LWRs are able to do this since despite a widespread belief that all nuclear plants are inflexible, they are both capable of doing so and are already providing the required services (Pouret et al., 2009). The ways in which each reactor type achieves this are slightly different, which will be discussed below.