In recent years, worries about runaway climate change forced by rising levels of carbon dioxide in the atmosphere have led many authorities to set ambitious targets in terms of reductions in future carbon dioxide emissions. The EU’s main instrument for implementing its climate policy is its Emissions Trading Scheme (ETS). The ETS provides an incentive to reduce carbon emissions and, for new entrants to the market, the carbon price is a direct tax on all emissions.

One of the difficulties of the scheme is that, in an economic downturn, there is likely to be a general excess of carbon allowances resulting in a fall in the ‘carbon price’. In February 2009, for example, carbon allowances fell to a little over €10 (US $13[10]) per tonne (of CO2) compared to €30 ($40) six months previously. In consequence the UK Government has recently proposed12 a carbon price ‘floor’ as part of its strategy for secure, affordable and low-carbon electricity. This is to be applied in the UK at a rate of £16 ($25) per tonne in 2013, rising to £30 ($46) in 2020 and £70 ($108) in 2030. As we shall see, one effect of this is to make nuclear power and renewables more competitive by driving up the costs of their CO2-emitting competitors.

PCGE assumes a carbon price of $30 per tonne of CO2. This looks reasonable when compared with an overall average price of around €20 ($27) in the three year period ending November 2008. When considering the next two to seven decades, however, it seems more likely to be an underestimate for Europe although possibly not for other countries. For the base calculation, therefore, we assume a carbon price of $50 per tonne of CO2 (Table 5.2) but will also consider a carbon price of zero.

The CO2 emission figures given for gas in Table 5.2 reflect only the carbon dioxide coming from the burning of gas. They do not include so-called fugitive greenhouse gas emissions that arise during gas production. These include direct leakages of gas to the atmosphere from boreholes and flaring of unwanted gas. From data presented in WEO 2011,1 3 it is estimated that, if they were to be included under the ETS rules, they would increase equivalent CO2 emissions above those shown in Table 5.2 by around 18%. For unconventional gas, fugitive emissions are likely to be higher still because fracking (fracturing of underground sediments by hydraulic pressurisation), which is used to release gas from low permeability deposits, requires as many as ten times more boreholes than conventional extraction. If this were covered by the carbon trading rules it would make a significant difference to the carbon price and must, therefore, be considered a potential risk.

6.6.1 Uranium resources — long-term prospects

An orebody is, by definition, an occurrence of mineralisation from which the metal or mineral is economically recoverable. It is therefore relative to both costs of extraction and market prices. At present neither the oceans nor any granites are uranium orebodies, and they are unlikely to become so even if prices were to rise substantially (Table 6.2).

Measured resources of uranium, the amount known to be economically recoverable from orebodies, are thus also relative to costs and prices. They are also very dependent on the intensity of past exploration effort, and are basically a statement about what is known rather than what is there in the Earth’s crust — epistemology rather than geology.

From time to time concerns are raised that the known resources might be insufficient when judged as a multiple of present rate of use. But this is the Limits to Growth fallacy, a major intellectual blunder recycled from the 1970s, which misunderstands the meaning of resource data, taking no account of the very limited nature of the knowledge we have at any time of what is actually in the Earth’s crust. Our knowledge of geology is such that we can be confident that identified resources of metal minerals are a fraction of what is there.

Table 6.3 and Fig. 6.3 show the current known recoverable resources of uranium by country. Uranium is not a rare element and occurs in potentially recoverable concentrations in many types of geological settings. As with other minerals, investment in geological exploration generally results in increased known resources. Over 2005 and 2006 exploration effort resulted in the world’s known uranium resources increasing by 15% in those two years.

The most common uranium product from mines is U3O8, which contains about 85% uranium. Table 6.3 refers to pure uranium, but the production figures may be expressed in terms of U3O8 by multiplying by 1.1793.

Reasonably Assured Resources plus Inferred Resources, to US$ 130/kg U, 1/1/09, from OECD NEA & IAEA, Uranium 2009: Resources, Production and Demand (‘Red Book’).

The current global demand for uranium is about 68 500 tU/yr (Fig. 6.2). The vast majority is consumed by the power sector with a small amount also being used for medical and research purposes, and some for naval propulsion. In total this mined uranium accounts for 78% of annual nuclear power station requirements. The remainder is made up from secondary supplies as outlined.

6.3 Known uranium resources (000 tU). IAEA & NEA Red Book 2009, resources to $130/kg U.

Thus the world’s present measured resources of uranium (5.4 Mt) in the cost category a bit above present spot prices and used only in conventional reactors, are enough to last about 80 years. This represents a higher level of assured resources than is normal for most minerals. Further exploration and higher prices will certainly, on the basis of present geological knowledge, yield further resources as present ones are used up.

In the third uranium exploration cycle from 2003 to the end of 2009 about US$ 5.75 billion was spent on uranium exploration and deposit delineation in over 600 projects. In this period over 400 new junior companies were formed or changed their orientation to raise over US$ 2 billion for uranium exploration. About 60% of this was spent on better defining and quantifying previously known deposits. All this was in response to the increased uranium price in the market.

The price of a mineral commodity also directly determines the amount of known resources that are economically extractable. On the basis of analogies with other metal minerals, a doubling of price from present levels could be expected to create about a tenfold increase in measured economic resources, over time, due both to increased exploration and the reclassification of resources regarding what is economically recoverable. Thus, any predictions of the future availability of any mineral, including uranium, which are based on current cost and price data and current geological knowledge are likely to be extremely conservative.

This is in fact suggested in the IAEA-NEA figures if those covering estimates of all conventional resources are considered — another 5.5 Mt (beyond the 5.4 Mt known economic resources), which takes us to 160 years’ supply at today’s rate of consumption. This still ignores the technological factor mentioned below. It also omits unconventional resources such as phosphate/phosphorite deposits (up to 22 Mt U recoverable as by-product) and seawater (up to 4000 Mt), though this would be uneconomic to extract in the foreseeable future.

It is clear from Fig. 6.4 that known uranium resources have increased almost threefold since 1975, in line with expenditure on uranium exploration. (The decrease in the decade 1983-1993 was due to some countries tightening their criteria for reporting. If this were carried back two decades, the lines would fit even more closely. The change from 2007 to 2009 is due to reclassifying resources into higher-cost categories.) Increased exploration expenditure in the future is likely to result in a corresponding increase in known resources, even as inflation increases costs of recovery and hence tends to decrease the figures in each cost category.

About 20% of US uranium came from central Florida’s phosphate deposits to the mid 1990s, as a by-product, but it then became uneconomic. With higher uranium prices today the Florida resource is being examined again, as is another lower-grade one in Morocco.

A technological factor also bears upon consideration of future uranium supplies. Widespread use of the fast breeder reactor could increase the utilisation of uranium

50-fold or more. This type of reactor can be started up on plutonium derived from conventional reactors and operated in closed circuit with its reprocessing plant. Such a reactor, supplied with natural or depleted uranium for its ‘fertile blanket’, can be operated so that each tonne of ore yields 60 times more energy than in a conventional reactor. Furthermore, the 1.5 million tonnes of depleted uranium left over from enrichment plants around the world can be used a fuel for fast reactors.

There is no reason to anticipate any shortage of uranium that will prevent conventional nuclear power from playing an expanding role in providing the world’s energy needs for decades or even centuries to come. This does not even take into account improvements in nuclear power technology, which could effectively increase the available resource dramatically.

W. WIESENACK, InstituttforEnergiteknikk, Norway

Abstract: Fuel assemblies and their components are subjected to the most harsh conditions existing in a nuclear reactor. They are designed and manufactured to satisfy stringent functional and safety requirements for normal operation and transient conditions. With an emphasis on boiling and pressurised light water reactors and going from general principles to details, the chapter describes the main components of a fuel assembly and how design and functionality are related to operating conditions, safety criteria and reactor physics. Phenomena affecting fuel rod endurance are addressed and illustrated.

Key words: fuel assembly, fuel rod, pellet, cladding, water rod, spacer grid, debris filter, nozzle plate, tie plate, power distribution, fuel failure, safety criteria, fuel fabrication, cladding fabrication.

8.3 Introduction

A reactor core is composed of fuel assemblies, also called elements or bundles. The fuel assembly is the basic fuelling unit manufactured by fuel vendors, transported to the reactor site, inserted into and removed from the core, and stored on site or, later, reprocessed or disposed of to a waste repository. It consists of individual fuel rods arranged in a square (BWR/PWR), hexagonal (VVER) or circular matrix (CANDU reactor and AGR). The rods are kept in position by grid plates at the ends of the fuel assembly and by spacer grids in between. They are filled with cylindrical pellets consisting of fissile and fertile atoms (U, Pu, Th), most often as oxides.

Fuel assemblies and their components are subjected to the most harsh conditions existing in a reactor: high neutron and gamma flux, high temperatures in a corrosive environment, high pressure and thermo-mechanical loads. Their design and manufacture must satisfy a number of functional and safety requirements for normal operation and transient conditions and guarantee many years of reliable and economical in-core service. Fuel assemblies must: [18]

• be long lived without undue deterioration and thus permit high fuel burn-up, minimising handling and storage needs

• be suitable for intermediate and final storage or reprocessing

• allow removal of decay heat in accident conditions

Today’s fuel assemblies and the materials employed in them are the result of many years of development and efforts to perfect the fuel designs for the different types of water-cooled and moderated reactors. Their optimisation contributes to reducing electricity generating costs while adhering to high safety and reliability standards. Ideally, fuel should not fail under normal operation conditions, but a more realistic goal reachable with present products is less than one failed fuel rod in ten reactor operating years, which corresponds to about one failed fuel rod in a million.

Typical reliability issues leading to failure are debris fretting (caused by foreign objects stuck in the spacer grid) and pellet-clad interaction assisted by pellet chipping. Such failures are avoided, respectively, by enhanced spacer and debris filter designs and improved pellet production and quality control. Another reliability issue is the bowing of long structural parts in connection with long operation cycles and neutron flux gradients. Such effects may impede the insertion of control rods and influence the local thermo-hydraulic and nuclear conditions in a fuel element. Finally, phenomena associated with corrosion of the outer surface of the fuel rod have an impact on safety and reliability. Among them are hydrogen pick-up and embrittlement, the thermal resistance of the corrosion layer and oxide spalling leading to temperature differences and increased hydrogen concentration reducing the cladding ductility.

The fuel assembly design and the choice of materials are governed by safety criteria (NEA, 2003). The assembly as a whole must:

• be able to withstand the mechanical loads and accelerations occurring during transport and handling

• provide sufficient stability margins against buckling under axial loads and bear the hydraulic lift forces arising during normal operation and accident conditions

• accommodate differential axial expansion and stay sufficiently straight under the influence of neutron-induced material growth such that incomplete control rod insertion (IRI) is prevented

• ensure safe reactor shut-down for events as severe as loss-of-coolant accidents (LOCA) and reactivity insertion accidents (RIA)

The fuel rod component must:

• provide sufficient internal free volume to accommodate fission gases released during operation without exceeding the pressure limit defined by the associated safety criterion

• withstand the elastic and plastic strains due to pellet-clad mechanical interaction

• be sufficiently resistant against cladding corrosion and hydrogen pick-up for the envisaged discharge fuel burn-up

The typical components of a fuel assembly and their functions will be treated in detail in the next sections.

As has already been noted the cooling of the fuel is key to the preservation of the fuel-related containment barriers. Cooling the fuel reduces the chance of clad failure as well as maintaining the effectiveness of the fuel matrix as a means of confining fission products. Systems are provided to both maintain the coolant inventory as well as to ensure a continued heat rejection route. The safety systems providing the core cooling functions are described below for each reactor type.

PWR safety systems

At power the normal heat removal route is via the steam generators to generate steam to power the turbine. When the reactor trips the turbine is also tripped and the steam is diverted to the condenser. If it is not possible to reject heat to the condenser then energy is removed by discharging steam via the steam line atmospheric relief valves. The feed flow to the steam generators must be reduced to match the heat generation rate. In some plants the main feedwater pumps may continue to be used, but with reduced flow, while in others the main feedwater pumps are tripped and auxiliary/emergency feedwater pumps are brought into service. Even plants which have the capability to continue using their main feed pumps will have auxiliary feedwater pumps, which will be automatically initiated on detection of low water levels in the steam generator.

The auxiliary/emergency feedwater pumps can be either electrically driven or driven by small steam turbines. Given the importance of the feed systems it is normal to provide diverse as well as redundant feedwater trains. In many cases this is achieved by having both electric and steam-driven pumps, but some plants use additional electrically driven pumps with diverse electric supplies.

Provided that the circuit is intact there should be no loss of primary fluid. The plant will initially be maintained at ‘hot shutdown’ (subcritical at normal operating temperature and pressure), but if it is necessary to cool the plant down then it will be necessary to inject borated water to compensate for both the shrinkage of the coolant inventory and the reactivity increase associated with the cooling of the core. This is achieved via the CVCS system (see Fig. 10.8) but in some plants this may be supplemented by a diverse charging system drawing borated water from a dedicated high boron concentration water tank.

As with any high pressure system, the consequences of a failure of a section of high pressure pipework must be protected against. The classic design approach was to provide protection for the worst pipe break, often referred to as the ‘maximum credible accident’. For PWRs this is the failure of the hot or cold legs of one of the main circulation loops, which results in a large loss of coolant accident (LOCA). The high pressure coolant flashes off to a steam/water mixture as it is discharged into the containment, and because the whole circuit depressurises, the water in the vessel boils off, essentially emptying the circuit. This shuts down the fission process even without the control rods being inserted, though the protection system will also initiate a reactor trip.

The fission process may be shut down, but stored energy and decay heat must be removed so systems are provided to reflood and cool the core. This is achieved by the use of the emergency core cooling system (ECCS), which consists of a series of pumps, which draw borated water from the refuelling water storage tank (RWST) and pressurised tanks (accumulators) containing borated water. In the case of the limiting large LOCA, the accumulators will discharge into the cold legs when the circuit pressure falls below the pressure of the nitrogen cover gas in the tanks (~4.5 MPa) because they are normally only isolated from the circuit by non-return valves. There is normally one tank for each loop and they are sized so that they will refill the lower plenum and downcomers following a large LOCA. The reflooding of the core is completed by water injected by the low head safety injection system (LHSIS).

Although the design basis for the systems is set by a hypothetical double-ended guillotine failure of either the hot or cold leg, this is an extremely unlikely event since this type of failure is rare and the breach opening would tend to be progressive rather than instantaneous. It does, however, provide a limiting case for the water delivery rate required of the LHSIS and hence the pump specification.

Water is initially drawn from the RWST. The boron levels are such that the core will remain subcritical even if the control rods have failed to insert. The water will quench and then cool the core to limit the maximum fuel rod temperature and thereby limit fission product release. Water will be discharged through the breach in the circuit and will collect in the containment sumps. When the level in the RWST has fallen to a low level the LHSIS pump suction is realigned to draw water from the containment recirculation sumps. In some modern plants and in advanced PWRs, the refuelling water is stored inside the containment in an internal refuelling water storage tank (IRWST). This is in the bottom of the containment building and acts also acts as the sump removing the need to switch over to recirculation. Initially the cooling is provided by the thermal capacity of the cold RWST water but in the longer term the LHSIS water is cooled by the residual heat removal system heat exchangers, which are cooled by the component cooling water system.

The ECCS must be able to deal with the complete range of possible LOCAs, ranging from small pipe failures to large breaks. In the case of the smaller breaks the circuit pressure will not fall as rapidly and will tend to stabilise at a pressure where the core decay heat is balanced by the heat losses from the circuit. The circuit heat losses will mainly consist of the energy flow through the breach and any heat removed by the steam generators. Heat removal via the steam generators will only take place when the primary circuit pressure remains above the secondary circuit pressure. This means that to cope with all possible sizes of circuit breach, it is necessary for the ECCS to be capable of injecting water at a wide range of pressures. However, the LHSIS pump characteristic is such that it is capable of delivering large quantities of water at low pressure but has a limited delivery pressure head and so cannot be used at high pressure. Thus a second system is generally provided to inject at high pressure: the high head safety injection system (HHSIS). In some cases this system is referred to as the intermediate pressure injection system since it is designed to deliver at pressures below full operating system pressure. The charging system (part of the CVCS shown in Fig 10.8) can provide injection at or above normal operating pressures and can deal with very small LOCAs. In some designs the charging and HHSIS functions are carried out using the same pumps.

In general ECCSs inject into the cold legs of the circulation loops since these feed into the downcomers and the inlet plenum. However, it may also be necessary to inject into the hot legs. Since boiling may occur in the core the dissolved boron in the coolant will concentrate in that region and periodic injection into the top of the core via the hot legs was introduced to mitigate the build-up of boron crystals in the upper parts of the core. Some designs also have lines, which allow direct injection into the reactor vessel.

Figure 10.14 shows a schematic of the Sizewell B safety injection systems. Also shown is the containment spray system. In many plants the spray and low head injection pumps can be realigned to do either LH injection or spray duty.

12.5.1 Introduction

This section of Chapter 12 looks at the design of the high-temperature gas — cooled reactor (either HTR or HTGR), which was perceived to be a potential successor to the advanced gas-cooled reactor. The HTR is significantly different to the AGR and the Magnox designs as is discussed in more detail below but the fundamental difference is that the HTR was designed to operate at higher temperatures than the AGR resulting in further increased thermal efficiency (in the region of 50%).

The HTGR design was first proposed in 1947. The prototype Peach Bottom reactor in the United States was the first HTGR to produce electricity from 1966 through 1974. Fort St Vrain in the States was one example of this design that operated as an HTGR from 1979 to 1989 though this is now decommissioned and no further HTGRs have been built there since.

Small scale HTGRs have also been built in Germany (the AVR and THTR-300, which was actually a commercial design), and currently exist in Japan and China. Two full-scale HTGRs, each with 100-195 MW of electrical output are under construction in China. More recently, this reactor design type has been substantially updated and is now proposed in a form known as the very high-temperature reactor in the United States.

The pebble bed modular reactor (PBMR) is a particular design of HTGR, which was under development by a South African company. The project was for the construction of a demonstration power plant at Koeberg, near Cape Town, although this has now been postponed indefinitely.

Commercial HTGRs have never built in the United Kingdom. There was a prototype, the Dragon reactor, built at the United Kingdom Atomic Energy Authority (UKAEA) site at Winfrith as a test bed for fuel and other materials. Operation of Dragon started in 1965 with a power output of 20 MW but it was shut down in 1976.

Development of the HTGR in the United Kingdom was terminated following a comprehensive review by the government, in the 1970s, of which thermal reactor strategy should be adopted for the future. As such, there is no detailed design information available. Consequently, the sections below describe the basic design principles of the HTGR.

Once the fuel-clad gap is closed, or nearly so, in oxide fuel pins with zirconium alloy cladding, chemical reactions can occur between the oxygen that has migrated towards the pellet outer radius (see 14.2.10) and the cladding material. This leads to fuel-cladding bonding. The bond is primarily composed of zirconium oxide, but there are also some fission product compounds and (given a tin-bearing zirconium alloy) tin oxide present. A fuel-cladding bond also forms in fast reactor oxide fuel, but in this case the bond (often known as JOG from the French term joint oxyde-gaine) is primarily composed of fission products.

Stress-corrosion cracking (SCC) is an important cladding failure mechanism for zirconium alloy clad fuel during power transients in which strong pellet­cladding mechanical interaction occurs. The necessary requirements for failure are a sufficiently high clad stress, a sufficiently high concentration of corrosive fission product species at the clad inner wall, and sufficient time for crack initiation and crack propagation. Experimental results point to free iodine being the primary corrosive species. Thermodynamic equilibrium considerations suggest that all released iodine would be in the form of caesium iodide, and therefore that there would be no free iodine available for corrosion, and no SCC. The fact that this is not observed shows that radiolysis is crucial for formation of free iodine, and hence for SCC. Stress-corrosion cracking of Zircaloy cladding materials was the subject of an IAEA coordinated research project. More details of this phenomenon can therefore be found in the project report (IAEA, 2000a).

The level of decontamination with regard to uranium and plutonium is a constraint linked to recycling. Required performances to get uranium and plutonium within the norms and standards can be measured through separation or decontamination factors.

The decontamination factor (DF) of an element is defined by the ratio of the specific activities before and after an operation:

DF = (Impurity/Product)m/(Impurity/Product)out The DF requirements are particularity strict, i. e. for f—y decontamination:

DF = 1.5 x 106

U

DF = 7 x 107

PU

These purification factors are much higher than in the ordinary chemical industry.

PUREX enables these specifications to be reached through the use of several extraction stages (generally two for uranium and two for plutonium).

Environmental radioactive releases

Authorized gas and liquid releases are set by the relevant authorities. The assessment of releases makes sense if it is reported against the amount of reprocessed fuel or in relation to the electrical energy produced by the reprocessed fuel expressed in MW. year.

Atmospheric releases primarily concern krypton-85, tritium, halogens and aerosols; for liquid releases it is tritium, ft-у emitters other than tritium, caesium and strontium, and a emitters.

I n this section, we will first give a summary of some fundamental features of transmutation. We will use very basic physics concepts in order to characterize the TRU properties that play a key role in the definition of any transmutation strategy. In the next section we will indicate how the application of these fundamental features leads to the selection of the most appropriate reactor concepts for transmutation and indicates the major issues and consequences for the fuel cycle.

As indicated above, to achieve transmutation, TRU-loaded fuel is irradiated in a neutron field. Several features characterize the transmutation potential of a specific neutron field for each TRU isotope:

Fission of isotope A should be favored against (n, y) and (n, xn) reactions.

Starting from isotope A, reactions giving rise to A+1, A+2, etc., should be minimized. In any event, the radioactive properties of isotopes A+1, A+2, etc., should be carefully investigated in terms of decay heat, neutron production, etc., in order to evaluate all consequences of transmutation.

The isotopes that successively lead towards full fission should, as much as possible, be ‘neutron producers’ rather than ‘neutron consumers’, in order to allow a viable core neutron balance.4

Where uranium and plutonium are removed for recycling, immobilisation of fission product raffinate, the liquid HLW that results from reprocessing, proceeds by calcining — boiling to dryness and then heating in an oxidising environment. The fission product oxides are then dissolved in molten borosilicate glass to produce vitrified HLW. This is poured into stainless steel canisters and then stored for decades to allow the rate of heat production to subside before deep disposal. The absence of plutonium eliminates an important source of heat but, nevertheless vitrified waste has a higher heat output per cubic metre than spent fuel because one vitrified waste canister contains the fission products from several fuel assemblies. Over the course of a few hundred years, however, the virtual absence of plutonumn in vitrified HLW ensures that its heat output eventually falls below that of the same volume of spent fuel.11

Spent nuclear fuel (SNF)

SNF continues to generate heat long after it has been removed from the reactor and, unless destined for reprocessing, freshly discharged fuel is held in water- filled pools at the reactor site for 5-10 years (Chapter 15). Subsequently, the fuel may be sent for longer-term wet or dry storage before disposal. Because of the impact of heat generation on disposal (not least on its cost), the aim will be to allow the heat generation rate to reach an acceptably low value; usually this will be less than 1 kW/tHM. In broad terms the amount of heat produced by uranium SNF is proportional to its burnup. As discharge burnup has been steadily increased over the past 40 years, so too has the cooling time needed to allow decay heat to reach levels deemed low enough for disposal. MOX fuel, which because of the greater quantities of plutonium, produces even higher levels of decay heat, requires even longer cooling times.12 In some cases these times may be so long that reprocessing is deemed a necessity.

When describing commercial risk it is convenient to consider its relevance to three aspects: (i) overnight costs, (ii) financing costs, both of which apply during the construction phase, and (iii) the business case once the plant is operating.

Overnight costs

The overnight cost data presented in PCGE suggest that uncertainties in cost estimates for nuclear plant are no bigger in percentage terms than those for gas or coal. When thought of in absolute terms, however, they are significantly greater for nuclear than they are for coal or gas. Consequently, nuclear is the most sensitive of the three to inaccuracies in estimates of overnight costs. This represents a risk that results in all kinds of sophisticated analyses being brought to the problem including probabilistic assessments to obtain more robust cost estimates. Ultimately, however, what is needed most of all is practical experience of NPP construction. The hope is that, as more of them are built and as designs become more standardised, estimates will become less subject to error and contingencies will be reduced.