Figure 10.6 shows the RPV and its internals. The vessel consists of a cylindrical shell with a hemispherical lower head. Inlet and outlet nozzles are set into the upper part of the vessel with a flange above these. A flanged hemispherical head

Control rod drive mechanism

Upper support plate

Internals support ledge

Core barrel Support column

Upper core plate Outlet nozzle

Baffle radial support Baffle

Core support columns Instrumentation thimble guides Radial support

Core support

closure assembly is connected to the vessel flange by a system of threaded studs and nuts. The integrity of the pressure vessel is vital since its failure cannot be tolerated because it will almost certainly lead to a large release of radioactivity. Older RPVs were constructed from plates, which were shaped and welded together but modern RPVs are made from cylindrical forgings. This obviates the need for vertical welds, which will be in the highest stress regions. This is particularly important in the cylindrical part of the vessel adjacent to the core since this will
be subject to the highest radiation levels. RPV materials are subject to embrittlement as a result of prolonged exposure to radiation and the weld material is more susceptible than the forgings themselves. For this reason some plants use a very large single forging for the core region.

The reactor vessel’s internal structures support the core, maintain fuel assembly alignment, limit fuel assembly movement and maintain proper alignment of the control rod guide tubes between fuel assembly and control rod drive. They also direct the flow of reactor coolant, provide gamma and neutron shielding, provide guides for in-core instrumentation between the reactor vessel bottom head and the fuel assemblies and support the surveillance specimen capsules. The internals are divided into the upper core support structure, which is removed for refuelling, and the lower core support structure, which can be removed for vessel inspection. The lower core support structure, which directly supports the core, consists of:

• the core barrel, a cylinder surrounding the core

• the thermal neutron shield

• the lower core plate and the lower core support, supporting the whole weight of the core

• the baffle assembly, inside the core barrel, which limits the by-pass flow of the core

• the secondary core support

The upper core support structure, which is located above the core, consists of:

• the upper support plate, directly above the fuel assemblies

• the control rod guide tubes and the thermocouple columns and their supports

The internals are supported on a ledge formed in the reactor vessel wall just below the vessel main flange. The neutron shield is provided to give additional shielding of the reactor vessel material in regions where the neutron flux might otherwise cause significant reduction in fracture toughness. Reactor vessel material samples are sited adjacent to the shielding pads to permit checks to be made on this property. Cooling water enters the vessel via the inlet nozzles and flows, down the annulus (downcomer) between the core barrel and the RPV, to the lower head. It then passes through the core removing heat. Coolant then flows from the upper regions through the outlet nozzles and into the steam generators.

Typical Magnox coolant compositions are 1-1.5% CO v/v, 25-45 volume parts per million (vpm) H2, 10 vpm CH4 and 1 vpm H2O, the balance being CO2. In very simple terms, the impurities protect the graphite against corrosion, but act as a source of carbonaceous deposition.

Coolant control is effected by drier units (which remove water, including a large amount of tritiated water) and recombination units, which remove hydrogen, methane and carbon monoxide by the addition of oxygen over a catalyst bed. Oldbury Power Station also has a liquefaction plant for coolant clean-up, which removes impurities by distillation (including rare gases in the coolant, which are subsequently released to the stack).

Air ingress during outages is inevitable, and removed by purging prior to return to power.

Graphite corrosion has been a problem on some reactors, leading to loss of strength and reduction in the moderating power. As a consequence, Oldbury Power Station has moved to using slightly enriched fuel, and at one time was considering the use of Magnox fuel: a small assembly of AGR-type fuel pins with steel cladding and enriched oxide pellets.

14.2.1 Fission-induced heating

As the reactor is brought to power at the start of irradiation, fission of the fissile isotope atoms in the fuel (primarily 235U for uranium fuels, 233U for thorium — uranium fuels, and 239Pu and 241Pu for uranium-plutonium or thorium-plutonium fuels) commences. The kinetic energy of the fission fragments is converted to thermal energy as the fragments come to rest in the fuel matrix as a result of inelastic scattering. This causes heating of the fuel bars (for metallic fuel employed in Magnox reactors and in some fast reactors) or fuel pellets (for ceramic fuel employed in all other cases). Thermal equilibrium (i. e. a steady-state condition) is rapidly established at any given reactor power level, such that the rate of heat generation by fission is balanced by the rate of heat dissipation to the reactor coolant. The heat flux from fuel pellets/bars to coolant via the cladding (also known as the sheath or can) is associated with a radial temperature gradient, with the maximum temperature occurring at the centre of the fuel pellets/bars.

The radial temperature distribution is complicated by the non-uniform generation of heat within the fuel pellets. In fast reactors the non-uniformity is minimal. However, in thermal reactors there is significant neutron flux depression as the thermalised neutrons diffuse from the moderator into the fuel pins, leading to preferential fissioning and heat generation in the outer regions of the fuel pellets/bars. In uranium-bearing fuels this is accentuated by the generation of fissile 239Pu from epithermal neutron capture in fertile 238U as irradiation proceeds (Carlsen and Sah, 1981).

Heat transfer is primarily by conduction through the fuel pellets/bars and cladding (and any pellet-cladding, or bar-cladding, gap), and by convection at the cladding-coolant interface. Radiation heat transfer from fuel pellets to cladding is also important at high fuel temperatures (such as occur in fast reactor fuel pellets, or at high power in LWR, AGR and CANDU fuel).

If the fuel becomes overly hot (which is generally only possible during severe accidents), the fuel pellets or bars, or even the cladding, can melt. In the case of metallic fuel, a phase change can also occur whereby there is a rapid and significant increase in fuel bar volume, which can rupture the cladding. For an unalloyed uranium fuel bar (as is effectively used in Magnox reactors) this (first) occurs at a temperature of 660 °C, when there is a transition from the alpha phase to the beta phase (Greenough and Murray, 1962).

The first dry storage systems were single purpose systems for storing of spent fuel, in most cases, away from the reactor (AFR). Typically, vaults, silos and non­transportable casks were used. In the last few decades, development has shifted towards multipurpose storage containers.

Vaults

Vaults consist of above or below ground reinforced-concrete buildings containing arrays of storage cavities suitable for containment of one or more fuel assemblies. Shielding is provided by the exterior structure. Heat removal is normally accomplished by forced or natural convection of air or gas (nitrogen or helium) over the exterior of the fuel-containing units or storage cavities and exhausting the air directly to the outside atmosphere or, where another gas is used, dissipating the heat via a secondary heat removal system.8 Typical features of vaults are their modularity, which facilitates incremental capacity extension, separated shielding and containment functions, capability for containment monitoring, and a vertical fuel-loading technology. Spent fuel is received either wet or dry to such a facility using transportation casks. Spent fuel is transferred to a metal-lined storage tube (single fuel element) or a cylinder (multi-element canister), which is housed within a concrete storage cavity in the vault structure. After transferring the fuel, the storage tubes are dried and sealed; they are usually backfilled with inert gas to prevent oxidation of the spent fuel in storage.

16.6.1 Recycling of uranium from reprocessing

Uranium from processing has generally a 235U nominal content superior to natural uranium (0.72%) so that the use of reprocessed uranium is a gain not only in mass but also as separative work units (SWU), which represent the energy required to increase the concentration of 235U above that of natural uranium. For reprocessed uranium to be used in the fuel assemblies that power a commercial nuclear reactor, it needs to be re-enriched (an LWR will generally need uranium with a 4-5% concentration of 235U). In addition to the 235 and 238 isotopes, reprocessed uranium also contains traces of uranium-236 and uranium-232. These two isotopes cause some difficulties:

• Uranium-236 is neutron-absorbing so that reprocessed uranium needs to be over-enriched in comparison with a fuel prepared with natural uranium.

• Uranium-232 has daughter products (lead-212 and thallium-208) that emit intense py activity. This may pose problems for radiation protection in the enrichment and fuel fabrication plants.

• In addition to these drawbacks, there is a particularly acute commercial difficulty with gaseous enrichment plants where a significant mass of uranium (known as ‘hold-up’) is always retained within the cascade (sequential enriching system). If natural uranium and uranium from reprocessing were to be treated alternately, it would be necessary to empty the cascade between production campaigns, which may be unacceptable economically.

• Co-precipitation of uranium and plutonium can be also used to produce a mixed oxide (MOX) fuel.

MA/Pu separation (possibly followed by further minor actinide separation, e. g. Cm removal) is mostly envisioned for dedicated transmuters (both ADS and low CR critical FRs) and ‘double strata’ scenarios (see Fig. 17.5). Where the objective is to use dedicated transmuters to phase out nuclear power and reduce the total TRU inventory (Fig. 17.6), grouped TRU separation (i. e. removal without separation into its constituent elements) is the most suitable approach. If, on the other hand, the objective is sustainability of nuclear power (Fig. 17.4) all options are available: grouped TRU separation (for homogeneous TRU recycling) or MA/ Pu separation (for heterogeneous TRU recycling) with the additional option of further MA separation.

Hydrometallurgical and pyrometallurgical chemical separation (partitioning) processes are under development in order to implement P&T strategies as indicated in the different scenarios described in Section 17.4. For both processes, a key issue is losses during reprocessing and re-fabrication, which must fall well below 1% and should probably approach 0.1%.

Many countries have, over the past four decades, developed hydrometallurgical processes to recover TRU elements so that, unlike the standard PUREX process, it is not only plutonium that is removed. Such processes would decrease the radiotoxic inventories of nuclear waste. While some of these processes have reached the stage of laboratory-scale demonstrations, none has ever been implemented at the industrial scale. Most of the partitioning strategies rely on a three-step approach:

• separation of U (and sometimes also Pu or/and Np) from spent fuel dissolution liquors

• actinide(III) + lanthanide(III) co-extraction

• actinide(III)/lanthanide(III) separation, which is the most difficult step because of the similar chemical properties of these element groups

The processes developed worldwide use a range of extraction systems. When considered in terms of process development (number of cycles, amount of secondary waste generated, scale-up of equipment, process control, robustness, safety analysis, etc.) or solvent management (treatment for recycling, by-product management, etc.) they are more or less suitable for industrial implementation. Nevertheless, all of these schemes require further development if not innovation. Even the most advanced processes may be in need of consolidation and optimization.58

Pyrochemistry has been studied for more than forty years, first for the reprocessing of fuels from molten salt reactors and then for the reprocessing of metallic fuels. These fuels were intended for the Integral Fast Reactor (IFR) proposed by the Argonne National Laboratory and used in practice in the EBR-II reactor at Idaho National Laboratory (INL), where the irradiated fuel of EBR-II has been processed using a pyrochemical facility.

Pyrochemical reprocessing has some advantages over hydrometallurgy since fuels foreseen for the new generation reactors would differ from present day commercial fuels and, in particular, would have lower solubility in acidic aqueous solutions. Other advantages are more compact equipment and the possibility of locating the reprocessing facilities close to the reactors, thus reducing considerably the transport of nuclear materials. In addition, the radiation stability of the salt in the pyrochemical process compared to the organic solvent in the hydro chemical process offers an important advantage (e. g. shorter cooling times), especially when dealing with highly active spent MA fuel. Until now, two pyrochemical processes have been developed to the pilot scale, both in chloride media, for the reprocessing and fabrication of oxide and metal fuels, respectively.

Pyrochemistry is a long-term objective and still requires much development and viability assessment before implementation at a larger scale. The research and development is thus often carried out in the framework of international collaboration. Specific issues that will need to be explored include corrosion of containment materials, online monitoring, the production of secondary waste and the motion of molten salts in pipes.

17.7 Conclusions

P&T has been historically associated with the waste minimization goal, and, in the last two decades, has been mostly explored in this context. Since about the year 2000, however, the Generation IV (Gen IV) initiative59 has, admitting some exceptions from a total consensus, defined a set of more general goals for future systems in four broad areas: a) sustainability (more efficient use of the available U resources and waste minimization); b) enhanced economics, c) enhanced safety and reliability and d) enhanced proliferation resistance and physical protection. The objectives of Gen IV do include P&T (waste minimization), consistent with sustainability and non-proliferation objectives and mostly associated with future reactor deployment.

Many recent studies have demonstrated that the impact of P&T on geological disposal concepts is significant even if not overwhelmingly high. 42 By reducing the decay heat of the waste, it becomes possible to utilize the repository volume more efficiently. As indicated in Section 17.5, deployment of P&T techniques reduces the thermal output of high-level waste by a factor of at least 3. This reduces the needed repository gallery length by the same factor and the repository footprint by up to a factor of at least 9.

The environmental impact of a deep repository is less affected by the use of P&T because the calculated doses to humans who might inhabit the land above the repository in the far future are mostly dominated by a few long-lived fission products for which no practical transmutation strategy is applicable. Nevertheless, by reducing the hazard associated with the emplaced materials, P&T would remove much of the uncertainty and (on the part of the general public) unease regarding the creation of a man-made Pu or U ‘ore body’ deep below the surface. This is especially the case for so-called intrusion scenarios that bring man in direct contact with the disposed waste. Other benefits arise from some less likely radionuclide release scenarios or unwanted events such as an increase in actinide mobility due to changes in geochemistry and (probably unlikely) nuclear criticalities due to actinide accumulation.

P&T will never remove the need for deep geological disposal but it has, nevertheless, the potential to significantly improve public perception of the ability to effectively manage radioactive wastes by largely reducing the TRU waste masses to be disposed and, consequently, to improve public acceptance. Both issues are important to the future sustainability of nuclear power.

While the physics of transmutation is well understood, major challenges are found in the fields of chemical separation, transmutation fuel development and impact on the fuel cycle. Regarding transmutation-dedicated reactors, key demonstrations are still expected, in particular in the field of external neutron source driven systems (as ADS), in order to prove their feasibility.

17.8 Acknowledgement

The author gratefully acknowledges the detailed, patient and appropriate comments of the editor of this book, which greatly contributed to the improvement of the present chapter.

Following the terrorist attacks in the early years of the twenty-first century, it was decided that security measures were required for all radioactive materials in transit and, in response, the IAEA developed relevant guidance (IAEA, 2008c). This document states the following objective:

Since transport occurs in the public domain and frequently involves intermodal transfers, it is a potentially vulnerable phase of domestic and international commerce. This guide is intended to facilitate a uniform and consistent approach to security.

The objective of this guide is to provide States with guidance in implement­ing, maintaining or enhancing a nuclear security regime to protect radioactive material (including nuclear material) while in transport against theft, sabotage or other malicious acts that could, if successful, have unacceptable radiological consequences. From a security point of view, a threshold is defined for determin­ing which packages or types of radioactive material need to be protected beyond prudent management practice. Minimizing the likelihood of theft or sabotage of radioactive material during transport is accomplished by a combination of measures to deter, detect, delay and respond to such acts. These measures are complemented by other measures to recover stolen material and mitigate pos­sible consequences, to further reduce the risks.

It establishes as its scope the following:

This guidance applies to the security of the international and domestic transport of all packages containing nuclear material as defined in the CPPNM and associated publications, and radioactive material that may pose a significant radiological hazard to individuals, society and the environment as a consequence of a malicious act.

The recommended actions for security during transport of all radioactive materials follows a graded approach and, generally, are similar to the provisions listed in detail above for nuclear material. Since many of the shipments associated with the nuclear fuel cycle will contain nuclear material, the focus here has been on the requirements from Nuclear Security Series No. 13 (IAEA, 2011). For those shipments that do not involve nuclear material, the shippers and carriers should consult Nuclear Security Series No. 9 (IAEA, 2008c) for further guidance.

Equation 5.3 makes it clear that a plant’s expenditure on fuel and carbon will be strongly influenced by its net thermal efficiency (i. e. thermal efficiency after allowing for on-site electricity usage). The thermal efficiency data provided by PCGE are not net values and the figures presented here in Table 5.2 are therefore set at about two percentage points lower than what is normally expected from modern plant.

As shown in the table, there is an expectation that the addition of carbon capture equipment will reduce the thermal efficiency of coal-fired plant from 40% to around 35%. Use of waste heat, as in combined heat and power schemes (CHP), is an obvious way to conserve energy and generate additional income; it does not, of course improve thermal efficiency in the sense of the efficiency of electrical conversion: the contrary, in fact. Such schemes are not included here.

The table provides a simple calculation of the additional cost of generation if CO2 emissions are priced at $50 per tonne. Coal with carbon capture is assumed to remove 90% of the CO2 so that the carbon cost is 10% of that without removal. The fact that CO2 releases are lower from natural gas than from coal is a result of the higher calorific value of gas (when expressed in terms of the mass of fuel) and the greater thermal efficiency of CCGT.

The main immediate trend is the expansion of world nuclear power generation capacity using current reactor and fuel cycle technology. This will increase uranium demand from today’s 63 000 tU/yr to about 107 000 tU in 2030 (WNA 2011 reference scenario — upper is 137 00 tU, lower is 52 000 tU).

But by 2030 Generation IV reactors are expected to be coming into service, and many of these designs will be fast reactors, so requiring virtually no new uranium supply from mines. The focus will be on reprocessing used fuel and recycling it.

6.5 Sources of further information

WNA information papers, including:

Country papers plus: Uranium in Africa, Uranium in Central Asia

Supply of uranium http://www. world-nuclear. org/info/inf75.html

World uranium mining http://www. world-nuclear. org/info/inf23.html

In situ leach mining of uranium http://www. world-nuclear. org/info/inf27.html

Uranium from phosphate deposits http://www. world-nuclear. org/info/phosphates_

inf124.html

Uranium from rare earths deposits http://www. world-nuclear. org/info/uranium_ rare_earth_deposits_inf130.html

OECD NEA and IAEA 2010, Uranium 2009: Resources, Production and Demand (‘RedBook’), NEA Paris

World Nuclear Association 2011, The Global Nuclear Fuel Market: Supply and Demand 2011-2030, WNA London.

Hore-Lacy, I, 2012, Nuclear Energy in the 21st Century, 3rd edn, World Nuclear University, London

In this section, the common characteristics of BWR and PWR fuel assemblies as well as features particular to each type will be described. CANDU and VVER fuel designs are not treated in detail in this chapter. Useful information about these latter types was given by Boczar et al. (2003) and in Nuclear Engineering International (2010). See also Chapter 11 for more details on CANDU fuel design.

The CANDU fuel design is very different from that of LWR fuel. In CANDU reactor terminology, bundle is used instead of assembly, and ‘fuel element’ means ‘fuel rod’. The CANDU fuel bundle is short and has few different components. It is composed of 37 or 43 elements (rods) of length 48 cm held together by bundle end plates in a circular arrangement (three rings plus centre rod). Each element consists of uranium-dioxide pellets encased in a thin-walled zircaloy tube. CANDU fuel is designed for on-power refuelling.

VVER (Vodo-Vodyanoi Energetichesky Reaktor = Water-Water Energetic Reactor) is the Russian type of pressurised water reactor. Its fuel assemblies differ in many respects from Western PWR assembly designs. Immediately noticeable are the hexagonal outer shape of the assembly and the corresponding triangular matrix of the fuel rods filling the space within a hexagonal shroud in the VVER — 440 type (VVER-1000 assemblies are without a shroud). The fuel rods are smaller in diameter than in PWRs, and the pellets have a centre hole. In contrast to PWRs, a VVER fuel assembly does not include space (tubes) for insertion of control rods.

Typical BWR and PWR fuel assemblies are shown in Fig. 9.1, Fig. 9.2 and Fig. 9.3, and typical dimensions are listed in Table 9.1 , The assemblies are approximately 3.9-4.8 m long and have a square cross section. Their bearing structure consists of end fittings linked by tie rods (BWR) or guide tubes (PWR). The fuel rods are inserted into this structure in a square lattice (14 x 14 to 18 * 18 for PWRs; 8 x 8 to 10 x 10 for BWRs) and kept from touching each other by spacer grids distributed along the length of the fuel rods. The BWR assembly has a shroud forming a flow channel to direct the flow of coolant along the fuel rods. Control rods can be inserted into the guide tubes of a PWR assembly, while the smaller BWR assemblies are arranged four and four in cells with space for insertion of a cruciform control blade unit between them.

The top and bottom end fittings are called ‘nozzle plates’ (PWR) or ‘tie plates’ (BWR). They provide lateral support to the ends of the fuel rods and control of the coolant flow through the assembly. The top nozzle or tie plate also prevents an

Cross wing

Bottom support

Screw

Transition piece

9.1 I llustration of a BWR assembly (SVEA-96 Optima3) and its main components (courtesy of Westinghouse Electric Company LLC).

upward movement of the fuel rods. It can be removed on site for retrieval of rods, inspection and repair. The BWR assembly has a handle attached to the top tie plate for lifting the assembly into and out of the core.

PWR assemblies are installed between the lower and upper core plate. Alignment holes at two diagonally opposite corners of both end fittings match pins in the upper and lower core plate. In this way, the fuel assembly is secured laterally in the reactor core. Axially, a PWR assembly is fixed by four strong,

9.2 Atrium 10xp BWR assembly (courtesy of AREVA pic).

9.3 PWR fuel assembly (courtesy of Westinghouse Electric Company LLC).

Table 9.1 Typical overall dimensions of PWR and BWR fuel assemblies Feature PWR BWR 14 x 14 15 x 15 16 x 16 17 x 17 18 x 18 9×9 10 x 10 Assembly length, (mm) 3900-4060 4060-4200 4060-4800 4060-4800 4830 4470 4420-4480 Assembly square width, (mm) 197-206 214-215 197-230 214 230 139 139 Rod length, (mm) 3730-3870 3860-3920 3880-4490 3850-4490 4390-4430 4075-4090 3890-4150 Number of fuel rods 176-179 204-208 236 264 300 72 91-96 Average heat rating, (W/cm) 204-220 203-238 176-211 171-200 166-167 158-160 124-158

multi-leaf springs, which are attached to the top end fitting plate by hold-down bolts. Other designs use four helical springs. The springs, when pressed down by the upper core plate, provide a counter force against the lifting force exerted by the upwards coolant flow.

The bottom nozzle or tie plate has several functions. It rests on the lower core plate carrying the load of the assembly, secures it laterally and directs the coolant flow to the assembly. It consists of a perforated plate preventing a downward movement of the fuel rods from the fuel assembly. Another important function is keeping foreign objects (for example metal particles, chips, turnings; collectively called debris) from entering the assembly and causing damage to the fuel rods (debris fretting is the main cause of fuel failure). To this end, the holes in the plate must be small enough to preclude or minimise the passing of debris while still allowing sufficient coolant flow. Fuel vendors have extended considerable effort to improve the debris filtering capability and developed various solutions, for example a bent flow path such that long, thin objects (wires) will not be able to pass through (Gotoh et al., 1999).

PWR assemblies contain 16-24 guide tubes made of a zirconium alloy. They provide the structural connection between the nozzle plates, serve as attachment points for the spacer grids and guide the control rods into and out of the assembly. The guide tubes are designed to ensure a fast drop of the control rods without damage to the latter or the assembly. The control rods fall easily through the upper part with a larger diameter and holes in the guide tube wall such that the water can be displaced without too much resistance. The lower part has a smaller diameter and no holes, thus serving as a dashpot softening the impact.

The centre position of the PWR assembly lattice is taken by a tube for insertion of in-core instrumentation.

A BWR fuel assembly is encased in a thin-walled square tube forming a flow channel. This feature is necessary to avoid reactor instability and cross flow of coolant and steam with the risk of locally inadequate cooling. So-called ‘flow trippers’ redirect the coolant onto the fuel rods for improved thermal performance. The tube is kept centred by leaf springs attached to the lower tie plate and is fixed to the upper tie plate by a fastener with a two-leaf positioning spring. The assembly casing also guides the control blade cross that can be inserted between four assemblies.

In light water reactors, the water has a double function as coolant and moderator. However, in a BWR, the moderator capability is diminished as more and more of the fuel channel towards the top is filled with steam. Water rods are therefore included in BWR assemblies to enhance neutron moderation especially in the upper part with most steam. In some designs, the water rods are cross shaped dividing the assembly into four sub-assemblies. The water rods may also function as tie rods connecting the lower and upper tie plate, and they may be used to attach the spacer grids.

The thin fuel rods, about 4 m long, would vibrate, bend and touch each other without additional lateral support between the end fittings. Spacer grids are therefore inserted about 50 cm apart and attached to the tie rods or guide tubes.

The grids usually consist of square cells formed by a lattice of metal straps. They support the fuel rods at several contact points, which are fixed, rigid dimples and springs providing lateral and axial forces to keep the rod in place while allowing for fuel rod thermal expansion and irradiation-induced growth.

The grids are mostly fabricated from zirconium alloy materials (low neutron absorption), but in some designs Inconel is used for the grids at both ends. Another variant is the bi-metallic grid where the dimples and springs are made of Inconel while a zirconium alloy is used for the straps. The middle grids may contain mixing vanes to increase the turbulence of the flow and coolant mixing within an assembly for improved thermal performance.

The fuel rods in all types of water moderated reactors (CANDU, VVER, BWR, PWR) consist of a cylindrical, zirconium alloy tube (cladding) filled with cylindrical fuel pellets, and two end plugs welded onto the cladding. The rod is filled with helium for good heat transfer through the gap between the pellets and the cladding. The cladding outside diameter is largest for CANDU elements (about 13 mm) and smallest for VVER reactors (9.1 mm), while BWR and PWR dimensions are in between. The rod diameters of the latter two types have decreased over the years as the assembly designs evolved from 7 x 7 to modern 10 x 10 lattices for BWRs and from 14 x 14 to 17 x 17 for PWRs. Typical values are 9.5 mm (PWR, 17 x 17) and 9.62-10.28 mm (BWR, 10 x 10). More details are given in Section 9.4 on fuel rod design and fabrication.