In computing LCOE, data were taken from PCGE for Europe and North America. Recognising that significant regional differences exist and that some costs will vary over time, this section examines the sensitivity of the LCOE to changes in six key parameters namely,

carbon cost; fuel cost;

O&M cost; delays and accelerations in construction;

discount rate; overnight cost.

Nuclear decommissioning costs are not included because they are relatively small. For each parameter we assume ±50% variation from the mean which, for those parameters for which data are provided by PCGE, generally represents a variation of between one and two standard deviations. An exception is the discount rate where we take values of 5 and 10%. The calculations do not consider the effect of availability because, with the exception of wind, the variability is small. However, a 50% increase in availability, which is feasible for wind, has exactly the same effect on LCOE as a 50% decrease in overnight costs.

The results are presented as tornado plots in Fig. 5.3. What this shows is that wind, nuclear and coal+CC are most affected by changes in overnight cost and discount rates so that, while nuclear is marginally less expensive than gas at 7.5%

discount rate, it is significantly more expensive than gas at 10% discount rate. Wind becomes competitive with gas and nuclear when overnight costs are reduced by half or, analogously, availability is increased to around 40%. For nuclear, the strong overnight cost component of LCOE indicates that it will also be sensitive to the installation of first-of-a-kind (FOAK) plant where overnight costs are higher. The low LCOE values that are obtained for nuclear when overnight costs are reduced by half explains why development effort is so often directed towards

її і і її і і n

Coal + CC

Coal

Carbon / / /

Fuel і і i

O&M

Dclay/Ac cirri Discount rate Overnight

60 70 80 90 100 110 120 130 140

LCOE $/MWh(e)

5.3 Tornado plots showing LCOE changes with (reading down for each technology) 50% changes in carbon price, fuel cost, O&M cost, delays/acceleration, discount rate (with LCOE values at 5 and 10%) and overnight costs. Central values are the same as Table 5.3.

this area and why countries such as South Korea, Russia and China, which cite low overnight costs, may attract interest from around the world.

Gas and coal are most affected by fuel and carbon costs (and, therefore, thermal efficiency also). Countries where natural gas is plentiful and cheap will clearly find it difficult to make an economic case for other forms of electricity generation although strategic issues might override this. Comparing gas and coal+CC we see that in regions where low cost coal is available (e. g. Australia and South Africa) carbon capture may be a competitive option provided that they do not also have access to cheap gas. Least affected by fuel and carbon costs are wind (which has no sensitivity at all) and nuclear.

What the study clearly shows is that carbon pricing is double-edged: on the one hand it penalises technologies with high carbon emissions but, because we shall continue to rely on fossil-fuelled generation for many years, it inevitably raises electricity prices overall.

The chemical and physical behaviour of different isotopes of the same element are virtually identical; however, there are very small differences that can be exploited to allow artificial concentration of a particular isotope. The binding forces between atoms or molecules are stronger in the solid and liquid phases than in the gaseous phase, so that the gas phase often offers the greatest potential for separation, based on the physical properties of the individual molecules.

Uranium metal and most uranium compounds have very high boiling points, making any form of gas phase processing very difficult. The exception to this is UF6 , which is a volatile solid at room temperature, having a vapour pressure of 10.6 kPa at 20 °C and a sublimation point of 56 °C. The critical point occurs at 64 °C and 151.7 kPa, meaning that liquid UF6 is not formed at atmospheric pressure. Figure 7.1 shows a phase diagram for UF6. The high vapour pressure of the material means that it can be handled as a gas at low pressure and room temperature, or at slightly elevated temperatures. Furthermore, fluorine has only one naturally occurring isotope, with a relative atomic mass of 19.00, so that any difference in the behaviour of 235UF6 and 238UF6 is due solely to the difference in the uranium isotope.

While UF6 exhibits a number of properties that make it well suited for use in a physical separation process, there are also properties that make it difficult to

handle. It reacts rapidly with water to form uranyl fluoride (UO2F2) and hydrogen fluoride (HF) via the reaction:

UF6 + 2H2O ^ UO2F2 + 4HF [7.1]

If the water is present as a vapour then the HF will tend to form as a gas whereas in bulk water the HF will form in solution, as hydrofluoric acid. UF6 may also react with organic materials, including hydrocarbon oils, to release HF. Both HF gas and hydrofluoric acid are toxic and corrosive.

UF6 can also react with metals directly although the rate of reaction with non­reactive metals tends to be slow and some metals, such as nickel and high nickel alloys are effectively resistant. Other metals, such as stainless steel, carbon steel and aluminium may be used for construction under appropriate conditions and with an appropriate allowance made for corrosion. Hydrofluoric acid will be formed in the presence of water, which tends to be far more corrosive than dry UF6.

It is important to take appropriate safety precautions when working with UF6. It reacts rapidly with water in the atmosphere to generate HF, presenting a significant and immediate chemical hazard. The main hazard from the uranium in UF6 arises if it is absorbed into the body via internal exposure pathways (i. e. ingestion, inhalation or injury) with the chemotoxic hazard considered to outweigh the radiological hazard at low to moderate enrichment levels. Its volatility means that it poses a greater threat for internal exposure than most uranium compounds and also makes the spread of contamination more likely. Once enriched, the potential for accidental criticality must be considered, for the UF6 itself and for any reaction products and residues. The key factor in ensuring safety is the prevention of any means of uncontrolled release.

Despite the challenges noted above, UF6 is used in most enrichment processes and in particular those that have been commercially exploited. The process of manufacture is known as conversion.

The prediction and optimisation of assembly power distribution is a key input to a thermo-hydraulic evaluation and safety assessment. From a nuclear physics point of view, PWR assemblies are less complicated than BWR assemblies. The analysis of the latter has to deal with the use of control rods during core depletion producing axial neutron flux and power variations and with the formation of a void (steam) in the assembly channel. Whole core codes (3D) must therefore be linked to a thermo-hydraulic code to provide a realistic thermo­hydraulic feedback of neutron behaviour. Details such as the exact distribution of steam and liquid water can have a considerable influence on the results (Jatuff et al., 2006).

Figure 9.4 shows the relative power distribution in a 10 x10 array of fuel rods in a BWR-type assembly; the numbers indicate the percentage of average power. In contrast to the description in Section 9.2, the central water rod is omitted. The control blade cross is assumed to be withdrawn, and the power distribution is point-symmetric to the centre of the assembly. The rod power, obtained with the Helios lattice code (Wemple et al., 2008), is highest at the corners (139%) and decreases towards the centre where it is lowest (80%). The reason for this behaviour is the inhomogeneous distribution of moderator and fuel. Fission

neutrons are moderated not only within the assembly, but also in the water slab around it from where they diffuse to the fuel and cause new fissions. Due to self­shielding, fewer neutrons reach the inner rods, and power is lower there.

Such an uneven power distribution is not desirable because of non-optimal fuel utilisation and power limitations imposed by the coolability of the highest rated rods. As illustrated in Section 9.2, BWR assemblies have interior water rods or water channels to improve neutron moderation and thus to obtain a more even power distribution. The effect of a particular design (Westinghouse SVEA 96) is shown in Fig. 9.5. The internal water cross divides the assembly into four equal parts, and the lowest power (89%) now occurs in the interior of a sub-assembly. The highest power (124%) is still generated in the corner rods, but overall, the power distribution is more even than in the first case.

Real BWR and PWR assembly designs are even more complicated and try to decrease power differences by also varying the enrichment and by adding a burnable poison such as gadolinium to the fuel. This is illustrated in Fig. 9.6 (Brunzell, 2006). A good designs will work for the entire in-core service time of

235и

1.6 w/o

2.6 w/o

3.4 w/o

3.9 w/o

4.4 w/o

4.9 w/o

an assembly from 0 up to 60 MWd/kgU burn-up or even more. Optimal solutions can be found with modern core physics and computing tools as demonstrated by Martin-del-Campo et al. (2007).

The cooling and moderation in a BWR assembly suffers from increasing voidage towards the top of the assembly. A design attempting to counteract this effect employs part length rods extending from the bottom of the channel part way to the top of the channel into the boiling region. They improve the flow distribution in the upper part of the assembly by channelling steam and enabling a high steam — to-water slip ratio. This increases the density of the moderating water around the remaining rods in the upper region of the assembly and improves the axial power distribution. The design also reduces the two-phase flow pressure drop in the upper part of the assembly and improves the core’s thermo-hydraulic stability.

There is more variability in the design of PWR containments because it was the responsibility of the architect engineer (AE) rather than the NSSS vendor and in the US there were a number of different AEs. However, the general characteristics are similar in that the containments consist of either a strong steel-lined reinforced or pre-stressed concrete shell, or a strong steel shell (with a reinforced concrete building surrounding this). These provide both strength and leak tightness. A typical PWR large dry containment is shown in Fig. 10.17.

The pressure within the containment is generally reduced below atmospheric to inhibit leakage but one subset of dry containments uses this as a design feature. The sub-atmospheric containment is maintained at a negative pressure (~ 5 psi or 35 kPa) with respect to the outside atmosphere. This negative pressure provides some additional margin for response to design basis accidents, and therefore the design pressure and/or volume can be reduced accordingly.

A variation adopted by Framatome for the French 1300-MW plants was the use of two unlined concrete structures. The inner one is pre-stressed and provides the pressure-retaining capability while the outer reinforced concrete structure provides protection from external hazards. Leakage from the unlined inner structure is reduced by painting the walls with an epoxy paint and the interspace between the two containments is maintained at a reduced pressure and is filtered.

Although the containments are essentially passive in the short term they are provided with systems to remove heat and fission products from their atmospheres. This is normally done using spray systems (see for example Fig. 10.14). The spray water will both remove heat and help to remove fission products. To maintain fission product iodine in a non-volatile form the spray water is dosed to make it alkaline by adding chemicals either directly to the spray water or to the recirculation sumps. Some plants use fan coolers to remove heat from the containment, either as an alternative to the use of containment sprays or as a diverse heat removal route. The heat is rejected by passing air through a cooling matrix, cooled by the service water system.

The sustainability of nuclear energy operations is partly determined by the choice of nuclear fuel cycle. The commonly used once-through cycle is the most wasteful, with less than 1% of the extracted U being converted into energy. This increases the use of U as a fuel source. According to recent estimates, the existing and estimated additional U availability is sufficient to support a once-through cycle until at least the end of the current century (Vezzoni, 2011). After the end of this century, unless new mining or extraction techniques are developed, U availability will be the limiting factor. The most recent available assessments (NEA and IAEA, 2010) have fixed a maximum limit value for the identified resources (IR) available at a cost of less than 130 $/kgn of about 16 Mton. This quantity can be augmented by adding the uranium dispersed in phosphates (about 22 Mton (Pool, 1994)), which gives a maximum limit of 38 Mton. This estimate is subject to change, depending on improved exploitation techniques at higher U prices (as has occurred with oil (Vezzoni, 2011)). As an example, it is possible to add a new resource category (available at a cost of less than 260 $/kgn) in response both to overall U market price increase and increased mining costs that enables a 15% extension of the (conventional) resources available (NEA and IAEA, 2010).

In the short term, three main options have been investigated to take account of finite U resources (Vezzoni, 2011):

• the adoption of reactor cores with a high conversion ratio (CR), defined as the ratio of TRU fissile production/TRU destruction

• the adoption of very high burn-up fuel

• the recycling of plutonium in LWRs by the use of mixed oxide (MOX) fuel

A key development is the transition from the open fuel cycle, where all the spent fuel is disposed of without recovering Pu (and MAs), to more optimized closed or partially closed fuel cycles based on partitioning and transmutation (P&T) (Vezzoni, 2011). Within a closed fuel cycle, fuel can be recycled, waste reduced and partitioned (e. g. MAs separated) so that each fraction can then be dealt with more effectively. Progressive waste management techniques include the transmutation of selected nuclides, cost-effective decay-heat management, flexible interim storage and customized waste forms for specific geologic repository environments. Because most of the heavy long-lived radioactive elements are removed, such methods significantly reduce the toxicity and decay heat of waste heading for geological repositories. This makes it easier to store and dispose of these radioactive wastes. The Generation IV roadmap fuel cycle crosscut group (FCCG) found that an important limiting factor facing the once-through cycle is the global availability of repository space, particularly as new repository capacity will be needed in a matter of decades. Closed fuel cycles have the potential to reduce the pressure on repository space and performance requirements.

There are number of challenges still to be overcome, including further development of separation technologies and the feasibility of advanced burners (Vezzoni, 2011). Another potential problem is the partitioning and transmutation of fissionable materials, which could be seen as adding to the risk of nuclear proliferation. Advanced partitioning technologies for Generation IV systems are better since they aim to prevent the separation of Pu from other actinides, and incorporate features that reduce the accessibility and possibility of creating weapons from waste materials. The use of fast-spectrum reactors and repeated recycling may make it possible to lessen the radiotoxicity of all wastes to the point where the confinement requirements can be reduced to less than 1000 years. However, realizing this goal would require substantial further research into fuel recycling techniques.

One important recent development has been the design of highly durable ceramics for the immobilization and possible disposal of MAs as well as Pu from dismantled nuclear weapons. Several possible hosts for actinides have been investigated, including complex oxides, silicates, and phosphates (Ewing, 2007). The most studied phase (Ewing et al., 2004) is pyrochlore (A2B2O7 where A and B are generally rare-earth or transition metal elements) because of its:

• ability to incorporate actinides

• chemical durability

• resistance to radiation damage (at least for some compositions)

It has been shown that compositions can be adjusted so that the dose at which the material becomes amorphous due to alpha-decay damage can be substantially reduced (Ewing, 2008), as shown in Fig. 13.1. Such investigations can be

13.1 Predicted temperature dependence of amorphization in pyrochlore-related phases containing 239Pu (Ewing et al., 2004).

considered the initial stage in the design of waste forms for particular waste stream compositions and repository conditions. For example, in the future it could be possible to choose the waste loading of a material based on the interplay between radiation damage accumulation and the anticipated thermal future of the repository. The nuclear fuel cycle will therefore become safer with the development of highly durable materials for the ‘back-end’ cycle (Ewing, 2008).

In reactors employing daily load following, or other types of power cycling operation, the cladding accumulates fatigue damage. The accumulated damage, and hence the fatigue life, is dependent upon the cladding type, the number of cycles and the amplitude of the clad stress variations during each cycle. Further information on fatigue of zirconium alloy cladding can be found elsewhere (O’Donnell and Langer, 1964).

Cladding is also subject to fretting (wearing away of the cladding material by repeated relative motion of the cladding and material in contact with it). Grid-to-pin fretting wear may occur if flow-induced assembly vibration leads to relative motion between the grid straps and the fuel pins (although the aim is to eliminate this possibility by good fuel assembly design, i. e. by maintaining sufficient forces between grid straps and fuel pins by use of grid springs and dimples, and by optimising grid profiles to minimise vibrations). Debris-induced fretting wear of cladding is also possible. However, fuel assemblies are usually designed in such a way to prevent this, or at least to minimise the possibility of its occurrence. If fretting wear becomes too great the cladding can fail.

Release of liquid and gaseous effluents leads to their dispersion in the environment. For both types of release the main impact on man is through ingestion: sea food in the case of liquid releases and products such as milk, meat and plants for atmospheric releases. Rare gases are not absorbed by plants or animals and have no significant impact. Aqueous and terrestrial environments are sampled extensively for monitoring purposes.

To give an idea of the magnitude impact, releases from La Hague (France) produce a dose to the most exposed person of less than 0.03 mSv/year. This may be compared with doses from exposure to natural radioactivity of 1.5-6 mSv/year (depending upon location) and the authorized level of 1 mSv/year.

Regulations can vary between countries: Japan sets restrictive limits at Rokkasho-Mura plant particularly with regards to liquid releases except tritium. The United Kingdom is more restrictive than France regarding the tritium content in liquid releases but requirements are less stringent for a emitters.

In fast neutron spectrum reactors, the recycling of plutonium and MAs in critical systems can be carried out in either a homogeneous or a heterogeneous mode.810 In the homogeneous recycling mode, all the TRU elements are recycled together (i. e. ‘grouped’ TRU) and, after mixing with other fissile material, form the standard fuel for the reactor. In heterogeneous mode, on the other hand, MAs are separated from plutonium and placed as targets in specific sub-assemblies and managed independently from the standard fuel, which contains plutonium. When comparing the two modes in terms of core and fuel cycle performance, key issues are a potential increase in the (already positive) sodium coolant void reactivity coefficient, which is associated with homogeneous TRU recycling in, e. g., a sodium-cooled fast reactor (which could limit the MA content of the fuel), and unwanted increases in decay heat and spontaneous neutron emission, which are of interest for the fuel cycle, especially when considering fuel fabrication (see below).

Recent studies (see e. g. Ref 11) have concluded that there are no theoretical ‘upper limits’ for the MA content of fast reactor fuel and that any potential issues for reactor safety can be managed through appropriate design measures. A recent JAEA study 1 2 calculated the different MA compositions that will arise in fast reactor fuel when, during a transition from a LWR fuel cycle to a FR fuel cycle, it is used for continuous homogeneous recycling of grouped TRU material: see

Fig. 17.3. It is observed that the MA content varies from 1 to 4 wt% corresponding to natural in-core equilibrium values. It is possible that the actual content could be higher if there is the need to deal with MA legacies from previous operation of LWRs. This is one advantage of partitioning the Pu and MAs as it allows the option to utilize and manage the two materials in any favorable ratio, even if they were to be subsequently recombined for use in homogeneous recycling. Such flexibility could be used to alleviate any potential safety issues arising from positive sodium void reactivity coefficients or from radiation doses during fabrication.

Two main considerations — scientific and social — govern the siting of a radioactive waste disposal facility and it is fair to say that, in the past 15-20 years the emphasis has shifted from the first to the second. Thus, if we look at two 1994 (though still current) IAEA publications 1 617 on the siting of near-surface and deep-disposal facilities, we find a site screening process that, for the most part, is technically driven. The first consideration is that the site should be stable — free of natural phenomena that might destroy the engineered and natural barriers that contain the radionuclides. Typical unwanted effects are volcanism, seismicity, tectonism and, for surface sites, flooding. Water is usually the main vector for radionuclide migration and so another important issue is to find a site with suitable hydrogeology. The list of pertinent site properties goes on to include geochemistry, erosion, proximity of natural resources, transport links, potential for urban sprawl and so on. Despite this somewhat prescriptive approach, both documents take care to emphasise that the search is not for the ‘best’ site, which can never be known, but merely one that is adequate.

1n practice, this technocratic, top-down approach has met with strong public opposition and many national site selection programmes have succumbed at the hands of well-organised local campaigns. The US Yucca Mountain Project, for example, inched forward year after year in the face of locally mounted legal obstacles only to fail in 2011 when all funding was removed and the Department of Energy (DOE) shut down the project. This fulfilled a pledge made by Barack Obama when (presumably) seeking to influence the voters of Nevada during his 2008 run for the US Presidency.18 The DOE subsequently said that a Yucca Mountain repository was not a workable option and there were better solutions.19 Since then, in a further development, the Nuclear Regulatory Commission has ruled that the DOE did not have the authority to withdraw a plan approved by Congress.18 All this only underlines the view of the US Government Accountability Office19 that:

First, social and political opposition to a permanent repository, not technical

issues, is the key obstacle.

In the UK, no less than three attempts to find radioactive waste disposal sites have ended in defeats for government policy: the HLW site investigation programme was cut short in 1981 following strong public opposition, the mid-1980’s search for a near-surface site was abandoned shortly before the 1987 General Election20 and the attempt to construct an underground Rock Characterisation Facility at Sellafield, widely viewed as a Trojan horse, was rejected at a local Public Inquiry.21 In the light of this history, the House of Lords Select Committee on Science and Technology22 concluded that:

Public acceptance of a national plan for the management of nuclear waste is

essential and it has to be achieved at the local level (i. e. close to potential

repository sites), as well as within the country as a whole.

A new UK policy came into being in June 200823 in which the Government opted for a ‘voluntarism and partnership approach’. Recognising that ‘a community which hosts a geological disposal facility for higher activity radioactive wastes will be volunteering an essential service to the nation’, an Engagement Package and a Community Benefits Package will form part of the quid pro quo. There has been one Expression of Interest to date that has come jointly from two communities close to the Sellafield site where around 60% of the UK’s radioactive waste is in storage. There is hope that further volunteers will come forward. If they don’t, then those Sellafield communities have two choices: keep the waste on the surface, building additional stores as they are needed, or allow it to be taken underground and accept the promised community benefits.

In contrast to the stories of failure, the Swedish radioactive waste management company, Svensk Karnbranslehantering AB (SKB), was established in the 1970s by the NPP owners and began a site selection programme in the early 1990s. Swedish national policy requires a volunteer approach. Provided the regulators are satisfied, local municipalities have the final say on whether a repository can be constructed in their locality or not. This approach obliges SKB to work hand-in­hand with those who have offered their ‘backyards’ as potential sites. Three communities, soon reduced to two, came forward. Each already hosts a nuclear power plant. In March 2011, SKB finally submitted an application for construction of a deep repository for spent nuclear fuel at Forsmark in the Osthammar Municipality, about 100 km north of Stockholm. Osthammar already hosts a nuclear power plant and a repository for LILW (also operated by SKB) at a depth of 100 m.

Radioactive waste is a burden on society at large, but the example of Sweden (and others) demonstrates that it is most effectively dealt with when ways are found of working with host communities to turn it to mutual political, social and economic advantage.

Calculation of the levelised cost of electricity generation (LCOE) shows that the economics of nuclear power are dominated by the high capital cost of nuclear power stations, which is itself strongly influenced by the cost of capital and, therefore, the applicable discount rate. Using European and North American prices, nuclear power appears to be competitive with other forms of electricity generation so long as the discount rate remains below about 10% and carbon dioxide emissions are penalised at the rate of a few tens of US dollars per tonne. Another important variable is the price of fossil fuel since this governs the competitiveness of coal — and gas-fired generation and, to a large degree, determines the price of electricity.

Biases in the standard LCOE calculation are discussed and it is concluded that the most important of these is the adoption of the same discount rate across all technologies. Recent work argues that the added socio-political and commercial risk associated with nuclear, especially for the first-of-a-kind technologies currently envisaged, should be recognised by the application of higher discount rates. Such arguments, which have force in the context of private capital, produce high LCOE values that can make nuclear power look uncompetitive. High discount rates arise from elevated levels of commercial risk, which underlines the importance of its management and mitigation.

An examination of NPP financing indicates the importance of effective project management and risk mitigation measures during construction. The latter include allocation of risk to those parties who are best able to control it, risk sharing through wider equity ownership and government support through loan guarantees and similar devices. Once built, the profitability of a nuclear power plant is ultimately determined by the price of electricity. Fossil fuel price is a key determinant of this but policy-related matters such as the rules for electricity trading and penalties on carbon dioxide emissions are also critical. Seen from the point of view of a potential private investor in nuclear power, the role of governments in providing the necessary long-term stability to the electricity and carbon markets cannot be over-emphasised.