House of Commons - Environmental Audit

Select Committee on Environmental Audit Sixth Report

The nuclear option

64. Various organisations argued in their submissions that a new generation of nuclear power stations was needed both to keep the lights on and to reduce carbon emissions. Indeed, it is largely because concerns on both these scores have recently increased that the issue of nuclear power is once again being vigorously debated—after having been consigned to the margins at the time of the Energy White Paper. In this part of the report, we examine various issues relating to the nuclear option—including issues of timing and sustainability—and in doing so we touch upon some financial aspects. However, we discuss the cost of nuclear power later in the report in the context of financial markets and investment risk.

Learning from the past?

65. The history of civil nuclear power in the UK over the last 50 years has been characterised by extensive government subsidies, time and cost overruns, and poor operational performance. While the first generation Magnox power stations built during the 1960s proved somewhat less problematic, the second generation Advanced Gas Cooled Reactors (AGRs) built from the 1960s to the 1980s were beset with difficulties. In the worst case, that of Dungeness B, it took 24 years from the start of construction to commercial operation, and the plant has only operated on average at 37% of its planned generating capacity since then. In operational terms, almost all UK nuclear reactors perform badly in international comparisons.[89] This is reflected in an overall average load factor for all nuclear plants of 71% in 2004.[90]

66. The poor performance of UK reactors may partly be explained by the fact that, in building the Magnox and AGR plants, the UK was pursuing a unique technological approach. However, even in the case of Sizewell B—the UK's only pressurised water reactor (PWR) which was built in the 1980s to an American design—the construction costs escalated from [£1.8] billion to over £3 billion, while generating costs have been estimated at over 6p per kWh—about twice the current cost of electricity from gas or coal.[91]

67. Some of the submissions we received argued that there had been a good track record of delivering nuclear power stations to time and cost in Asia (Japan, China, and South Korea), and that this could be replicated in the UK.[92] Apart from the difficulty of verifying these claims,[93] there are a number of other factors which make such a comparison less relevant. These countries do not operate the kind of liberalised market which exists in the UK, nor are the designs being built those which would be likely to be on offer in the West.[94] Indeed, many of the nuclear power plants being built elsewhere in the world are also subject to substantial overruns, and some may never be completed.[95] In their evidence to us, EDF pointed to the successful deployment of nuclear on a large scale in France, and we certainly agree that this does show what can be achieved. But EDF remains a state owned company, and there are significant questions about the amount of government subsidies provided to the nuclear industry over the last 40 years. Also, issues relating to waste disposal and decommissioning still have to be resolved.

68. The key issue at stake here is to what extent one can expect a new generation of nuclear power stations to be relatively trouble-free in terms of both construction and initial operation. Some of the memoranda we received were optimistic to the point of complacency in this respect. British Energy, for example, stated that a new programme "will not use a prototype technology—the 'winning' technologies have emerged (ie the AP 1000, EPR, and Candu designs)."[96] However, as many have pointed out, no western country has yet built a Generation III power station; and, while they are developed from existing designs, it is by no means certain that technological difficulties will not be encountered. For example the N1 Framatome design (on which the EPR is itself based) was supposed to incorporate lessons learned from the extensive French nuclear programme, but significant problems were experienced for the first four years of operation with load factors averaging only 40%.[97]

69. The joint memorandum submitted by DEFRA and the DTI recognised this issue in relation to the costs of nuclear power. "There is very little evidence on the robustness of cost estimates for new nuclear build as there is currently only one plant being built in Europe. […] Past experience of cost overruns in non-liberalised electricity markets, however provides no guide to the prospects for new nuclear build in a liberalised market."[98] Cost overruns are intimately connected with construction and operational delays, and it is therefore interesting to note that the EPR now being constructed in Finland—the first Generation III reactor to be built anywhere in the world—is already running over six months late, even though construction only began in mid-2005.[99] The suppliers are confident that they can make up for this delay, but it remains to be seen whether the reactor will in fact become operational in 2009.

70. The past history of the nuclear industry gives little confidence about the timescales and costs of new build. This does not mean that a new generation of nuclear power stations cannot be built to time and cost, but it does mean that investors have little basis for assessing the risks involved and may therefore require a higher rate of return.

Could nuclear be built in time?

71. A key argument put forward by the nuclear industry is that a new generation of nuclear power stations is vital in order to address the potential shortfall in generating capacity resulting from the decommissioning of old coal and nuclear plant. As we have set out above, some 20GW of new electricity generating capacity will be required by the end of 2015, and possibly as much again required by 2025.

72. All the evidence the Committee received indicated that a new series of nuclear power stations could not be built in time to address the need for new capacity by 2016 or even earlier. Those arguing against nuclear power tended to provide more pessimistic assessments such as the timescale for nuclear new build which Tom Burke set out in his evidence to us:

Government decision in 2008;

2 years to draw up detailed design specification (to 2010);

2 years for the Nuclear Installations Inspectorate to approve the design (2012);

3 years for a public inquiry resulting in construction start in 2015;

5 years construction (to 2020);

1 year operational testing (to 2021);

commencement of commercial generating (2021).[100]

The Secretary of State himself confirmed that it might currently take from 15 to 17 years before a new nuclear power station could become operational, though his official was quick to point out that this was "just the kind of issue which the [Energy] review needs to look at."[101]

73. A large part of the time required to build new plant relates to the design and planning stages, including the need to obtain the relevant consents. In the case of Sizewell B, for example, it took 7 years from the decision in principle to the start of construction, with the construction phase itself lasting a further 7 years. Attention has therefore focussed on reducing this timescale by streamlining the planning and consents process, and both BNFL and British Energy suggested in their memoranda that the Government needed to provide greater clarity here. After referring to the past problems of the UK civil nuclear programme, BNFL go on to state:

Today's deregulated market would bring a very different framework for building new power plants. In addition, the regulatory and approvals processes in place in the past, which allowed delays and re-design to become the norm, would act as a major deterrent to private sector investors. An improved delivery process would be required in future for a nuclear project to become a reality. This is achievable without legislative change but Government leadership is required to provide the necessary resolve. Such a process would retain the rigorous scrutiny and opportunity for democratic participation and challenge, but would have scope and timeframe clearly defined to bring predictability to the overall process. This means that key approvals need to be granted before construction with a well-defined scope and timetable for further approvals during construction and commissioning.[102]

74. It is not entirely clear what changes BNFL are asking for here, though they go on to suggest that the approvals process should be geared towards the fact that any future UK plant would be an internationally recognised standard design (eg the EPR and AP1000).[103] Yet, as Professor MacKerron pointed out, every country continues to feel the need to have its own safety regulatory system for nuclear power,[104] and—in the absence of far stronger international frameworks and standards—any move towards a reliance on approvals granted by another country would represent a momentous and unacceptable step. Nor would it be acceptable if there were to be any weakening of the regulatory consents process itself: in the case of nuclear safety issues are of overarching importance, and the past history of the industry in the UK provides various examples of both managerial and engineering failures.

75. One possible approach to shortening the consents process would be to ask the relevant authorities to 'pre-license' reactors before sites are selected. The Secretary of State told us that "As I understand it, you could get moving with pre-licensing pretty quickly and save about three years on the timescale if you were going down this route."[105] But the DTI were unable to tell us whether this would require legislative changes, and it pointed out that there was no clear definition of what pre-licensing meant.[106]

76. The Government has now commissioned from the Health and Safety Executive an expert review setting out the potential health and safety risks arising from recent and potential energy developments and on the HSE's approach to ensure that risks arising from these are sensibly managed by industry. While the ostensible purpose of the review may be to examine safety issues arising from increased gas storage, CCS technology, and increased renewables, the terms of reference specifically include "consideration of a new generation of nuclear power stations and in the event of nuclear build, the potential role of pre-licensing assessments of candidate designs." Indeed, in the letter commissioning the review, the Minister of State for Energy specifically wrote:

On nuclear, we would like an assessment from HSE of how they might go about the appraisal of reactor designs in advance of specific proposals for new build. This will be useful for the Review making a recommendation in taking the decision on whether there is a potential role for a new generation of nuclear power stations to help us meet our medium and long-term energy goals.[107]

77. BNFL also suggest that the timescale from decision in principle to the start of construction should be reduced significantly below 5 years.[108] But even 5 years seems ambitious, particularly when BNFL point out in the previous paragraph that many other proposals for CCGT or wind power have taken longer than this. Given the extent of public concern about nuclear, a pre-construction period of 7 years (as set out by Tom Burke) therefore seems more likely. As overt Government support, either through some form of statement or through a revised policy framework, is unlikely before 2007, it is difficult to see how even the first of a new series of power stations could become operational before 2019 at the very earliest. Various witnesses and organisations corroborated this timescale.[109]

78. The Government has committed itself to publishing a White Paper prior to any decision to promote new nuclear build. Even after any "signal" is given, a consortium would need to be formed, tendering undertaken, detailed design specifications drawn up, and various permissions obtained. It is difficult to see how the planning and consents process could be reduced below 5 years without a fundamental reform of regulatory and statutory procedures which might compromise both safety and public accountability. It is worth observing in this connection as an aside that planning procedures affect more than just nuclear, and that other technologies such as wind may also be disadvantaged. Indeed, several years ago the Government proposed a new approach for major infrastructure projects of national significance, but abandoned it in the face of widespread opposition. If the Government wishes to reform the planning system, it must ensure that it does so on a comprehensive and even-handed basis.

79. A further timing issue in relation to nuclear relates to the fact that a programme of eight or ten plants could not all be built at once. An important reason for building a series of reactors is to reduce the overall costs through scale economies and design improvements. The first plant in the series would be more expensive and the planning process significantly longer—as fundamental issues of safety would necessarily have to be addressed in a planning inquiry in a manner which would not apply to subsequent plants of the same design. In their proposals put forward in 2002 the nuclear industry suggested that a gap of 18 months between successive plants could be achieved. Yet this would mean that, for a series of 10 AP1000 reactors, it would take nearly 14 years before the last became operational. Not only would such a time-lag reduce the extent to which nuclear could contribute to the new investment required over the crucial period from 2020 to 2025, but it could also mean that the plants might become technologically outdated by the time the series is completed. Indeed, by 2030 it is likely that a far wider range of renewable technologies will have become more cost-competitive, and that there may be alternative and highly competitive forms of nuclear power available such as the pebble-bed modular reactor.

80. Nuclear can do nothing to fill the need for 20GW of new generating capacity which will arise by 2016, as it simply could not be built in time. The Secretary of State himself acknowledged that it might take 17 years before the first of a fleet of new nuclear power stations could become operational. Even if planning, licensing, and construction stages could be reduced to 10 years in total, the earliest possible date for the first of a series would be 2017—still too late to plug the immediate gap. For the period beyond 2017 nuclear could begin to make a contribution—though, given the fact that successive nuclear plants might only come on stream at perhaps 18 month intervals, it might not be until around 2030 that the full generating capacity of a nuclear programme would be available.

Uranium supplies

81. An important issue in respect of nuclear power is the availability of uranium supplies. While it is true that uranium is plentiful—it is found, for example, in extremely weak concentrations even in seawater—energy is required to extract and refine it. The ratio between the energy required for this purpose and the useful energy produced from the uranium is critical, and the availability of high quality uranium ores is therefore of crucial importance in assessing the longer term potential of nuclear power.

82. We received a variety of evidence on this score. BNFL suggested that some 11 million tonnes of uranium was economically recoverable and that this was sufficient to last for around 170 years at current consumption rates. They went on to claim that "Looking at it another way, this would be enough to provide a lifetime's fuel for all of today's nuclear reactors worldwide, plus all those which might be built as far ahead as 2050, even in a scenario where world nuclear capacity were to triple to 1200 GW by that date." [110] By contrast, other organisations expressed concerns about the uranium resources pointing out that current supplies might only be enough to last 50 years. Indeed, the Oxford Research Group (ORG) provided an interesting supplementary memorandum in which they analysed the BNFL claim of 170 years availability and argued that it misrepresented statements made by the OECD NEA and the International Atomic Energy Authority (IAEA).[111]

83. BNFL's claim seems also to be at odds with the 2004 OECD/IAEA "red book" on uranium resources, as the following passages show:

At the end of 2002, a total of 441 commercial nuclear reactors were operating with a net generating capacity of about 364 GWe requiring about 66,815 tU. By the year 2020, world nuclear capacity is projected to grow to between about 418 GWe net in the low demand case and 483 GWe net in the high demand case. Accordingly, world reactor-related uranium requirements are projected to rise to between about 73,495 tU and 86,070 tU by 2020.

At the end of 2002, world uranium production (36,042 tU) provided about 54% of world requirements (66,815 tU), with the remainder being met by secondary sources including civilian and military stockpiles, uranium reprocessing and re-enrichment of depleted uranium.

As currently projected, uranium production capabilities including existing, committed, planned and prospective production centres supported by Known Conventional Resources (RAR and EAR-I) recoverable at a cost of <USD 80/kgU cannot satisfy future world uranium requirements in either the low or high demand cases. Thus, secondary sources, ie excess commercial inventories, the expected delivery of Low Enriched Uranium (LEU) derived from Highly Enriched Uranium (HEU) warheads, re-enrichment of tails and spent fuel reprocessing, are necessary to ensure adequate supplies in the near-term.

However, secondary sources are expected to decline in importance, particularly after 2020, and reactor requirements will have to be increasingly met by the expansion of existing production capacity, together with the development of additional production centres or the introduction of alternate fuel cycles. However, significant and sustained near-term increases in uranium market prices will be needed to stimulate the timely development of resources. Because of the long lead-times necessary to discover new resources and bring them into production (typically in the order of 10 to 20 years or more), there exists the potential for the development of uranium supply shortfalls and significant upward pressure on uranium prices as secondary sources are exhausted.[112]

84. Put bluntly, uranium mines only supply just over half the current demand for uranium, and the situation is likely to become more acute as current secondary sources decline in importance. Indeed, the nuclear industry itself expressed considerable concerns about this issue at the World Nuclear Association's (WNA) annual conference in London in September 2005. A presentation of the findings contained in the latest WNA Market Report highlighted the fact that uranium supplies were expected to increase until about 2014, but would reduce slightly thereafter and then remain level until 2030. Moreover, while fuel availability will be sufficient to satisfy demand in the 'low-demand' scenario (a declining percentage of nuclear power globally), it would not be sufficient to satisfy demand in the 'reference' scenario (limited growth of nuclear), and would be far short in the 'high demand' scenario (significant global growth of nuclear). The presentation concluded:

….fuel supply is potentially short beyond 2015, unless the lower demand scenario occurs. … future uranium supply is now a big issue. Actually, the uranium market has been concerned about it for some time and accordingly, the price has been increasing for the last couple of years. … One of our concerns is that uncertainties about fuel security in the future may depress possible investors' confidence in the nuclear power industry. This could potentially delay or cancel the nuclear programmes, currently set. [113]

85. The scale of the problem is illustrated by van Leeuwen and Smith's calculations that a 'global growth of nuclear' scenario, as envisaged by the MIT and in which 1000 new reactors were built, would result in the existing known reserves of uranium being exhausted in only 14 years.[114] As a result of such concerns, the price of uranium has trebled in the last few years after two decades of very cheap supplies. It is, of course, possible that further increases in price will stimulate additional investment in production capacity, but the key point which these quotes bring out is that future supplies of uranium—even for only a moderate increase in demand—are anything but assured. While fuel prices are currently not a significant factor in the economics of nuclear power, this situation could change radically as the heavily subsidised secondary sources decline in importance. Moreover, the suggestion that price increases could stimulate a move to different fuel processing cycles (such as the use of MOX, or the development of fast breeder reactors) is doubtful. The MIT 2003 report on nuclear comprehensively showed that alternative fuel cycles are significantly more costly than a straight-through cycle.[115] Indeed, the UK's experience in reprocessing spent fuel at Sellafield has been disastrous, while the development of fast-breeder reactors has been beset with difficulties world-wide. Such suggestions can in any case only undermine the argument for investing now in the current designs on offer.

86. Very recently, the Sustainable Development Commission published a report which it commissioned from AEA Technology on uranium supplies as part of its review of nuclear power. We were interested to see that the report corroborated our own view about how serious an issue this is. Indeed, it not only confirms the limited nature of known and estimated reserves available, but also refers to the extent of market failure as reflected in the low price of uranium and the extent of government involvement worldwide in the uranium trade.[116] AEA's overall conclusion, however, is surprisingly upbeat: it argued in essence that, as OECD/IAEA 'red book' reserves had increased by 40% over the last 20 years, there was no reason to suppose that such a trend would not continue in future. We agree that this may indeed occur—but then again it may not. And in any event there are likely to be further large price rises.

87. It is also relevant to consider the sources of supply. While Canada and Australia are currently the largest sources of mined uranium, supply is increasingly dependent on other states including Russia, Kazakhstan, Uzbekistan, Namibia, and Niger.[117] We find a certain irony in the argument sometimes used to support nuclear new build that the UK should not become so dependent on foreign supplies of natural gas when it may become equally dependent on uranium imports from similar foreign sources. More generally, the situation with respect to uranium supplies mirrors to a surprising degree the current situation with regard to oil and gas reserves—not only in the limited nature of known and estimated reserves, but in the geographical displacement of those reserves.

88. One final aspect of this issue relates to the use of stockpiled military plutonium. In his evidence to us, Professor Sir David King put forward an interesting argument that nuclear power stations using mixed-oxide fuels (MOX) offer an effective way of getting rid of unwanted plutonium while addressing the security of supply issue: "This is a matter of interest to me as a government advisor, how do we treat the plutonium stockpile that we have? [...] Either we treat the plutonium as an energy source or we treat it as a major waste issue. I think processing it through a nuclear plant is by far the preferable way to go forward."[118] However, the price of MOX is currently higher than ordinary uranium, and it is unclear how this might change in future. Moreover, we remain concerned about the lack of transparency in the pricing of the plutonium and the possibility of hidden subsidies to promote its use.[119]

89. Uranium mines can only supply just over half the current demand for uranium, and the situation is likely to become more acute as secondary sources—such as military stockpiles from decommissioned weapons—decline in importance. Such concerns, which are shared by the nuclear industry itself, may depress investment in new nuclear capacity, while the possibility of further large rises in the price of uranium could significantly alter the economics of nuclear power and render it less attractive to investors.

Carbon emissions

90. Integrally linked to the availability of uranium supplies is the issue of the carbon emissions associated with nuclear. It is frequently said that nuclear is a zero-carbon source of power. While this may be true in terms of the operation of a nuclear power plant itself, significant energy is required for its construction and—more importantly—for the mining and fabrication of uranium to provide the necessary fuel, and the conditioning, transport, and final storage of radioactive waste.

91. The evidence which we received on this issue was limited. The nuclear industry cited a range of studies which all tended to show that life-cycle emissions from nuclear were among the lowest of any form of generation—lower, indeed, than wind. British Energy, for example, set out the results of a recent analysis of emissions from its Torness plant, which showed that emissions were just over 5 grams per kWh (g/kWh)—only slightly higher than the 3.3 g/kWh which a similar study carried out by the Swedish company Vattenfall carried out. Such figures contrast with emission levels of 800-950 g/kWh for coal plant and 350—400 g/kWh for gas.[120] BNFL referred to an IAEA study which suggested slightly higher levels (between 9 and 21 g/kWh).[121] Figures of around 5 g/kWh were cited by other organisations including EDF, the British Nuclear Energy Society, and the Nuclear Industry Association.[122]

92. By contrast, others argued that these studies understate the carbon emissions from nuclear and are hardly objective as they have been financed by the nuclear industry itself. Greenpeace cited the EU Externe study which suggested that emissions from nuclear, while still very low, were 50% more than from wind.[123] A recent paper from the Öko Institut, a leading German environmental and scientific research organisation, calculates nuclear emissions at 33 g/kWh and cited other international studies carried out by the IEA and CRIEPI which have suggested higher figures of between 30 and 60 g/kWh.[124] What is particularly interesting about the Öko paper is the comparative analysis of emissions across a range of technologies which demonstrates that gas ICE cogeneration is highly competitive with nuclear while biomass based cogen has far lower emissions. Their analysis is that nuclear is by no means the cheapest way of achieving carbon abatement.

93. Even higher figures of up to 120 g/kWh for carbon emissions from nuclear have been produced by van Leeuwen and Smith—about a fifth to a third of the emissions from a normal gas generating plant. They have also argued that, as good quality uranium reserves run out, a move towards lower quality uranium ores would significantly increase the energy required to refine it, and that the energy required can rapidly escalate to the point where more is needed than can be obtained from the use of the fuel. The nuclear industry has fiercely attacked the ongoing work of van Leeuwen and Smith, but the latter have issued detailed rejoinders.[125] The recent report of the Sustainable Development Commission on the extent to which nuclear can contribute to carbon reductions does not appear to have included consideration of this key issue—perhaps because of their conviction that further high quality uranium reserves would become available.[126]

94. At the very least, this controversy suggests that there is a need for far more objective analysis and discussion. Moreover, the EAC has commented in the past on the difficulty of carrying out thorough life-cycle analyses. The Secretary of State acknowledged that this issue needed to be examined and that it would be included in the Energy Review.[127] However, what is required here is not a simple desk-based review of existing studies, but an in-depth investigation which could hardly be accommodated within the time-frame set by the Government. Moreover, given the need for objectivity, it would not be appropriate for such a review to be conducted by the Government itself. What is required is an independent authoritative analysis such as that which the Royal Commission on Environmental Pollution could provide.

95. At present nuclear power can justifiably be regarded as a low-carbon source of electricity. However, the extent to which this can be sustained needs to be examined. There is some evidence to suggest that the level of emissions associated with nuclear might increase significantly as lower grades of ore are used. Given the concerns expressed by the nuclear industry itself over the adequacy of uranium supplies after 2015, we regard this as a serious issue and one which can hardly be resolved in the time-frame of the current Energy Review. In view of its importance, the Government should consider asking the Royal Commission on Environmental Pollution to report on carbon emissions associated with all generating technologies.

Safety, terrorism, and proliferation

96. One of the most emotive aspects of nuclear power relates to the potential impact if anything were to go wrong. The major incident at Three Mile Island and the disaster at Chernobyl have resulted in a heightened awareness of the possible environmental, social, and economic costs which could result from either an accident or an act of terrorism. Additional concerns arise because of the close links between the technologies involved in the civil and military uses of nuclear power, and the risk that a 'nuclear growth' scenario might well result in the proliferation of nuclear weapons worldwide.

97. Various organisations commented extensively and sometimes passionately on these issues.[128] Submissions from the nuclear industry tended to focus on the fact that a new generation of nuclear power stations would be inherently safer than existing designs, in terms of both operational reliability and the ability to withstand direct terrorist attacks. But they tended to dismiss other operational risks—such as the risks surrounding on-site storage of spent fuel and the increased transportation of nuclear material around the country. They also did not deal in any depth with the risks of proliferation in the event of a worldwide 'nuclear growth' scenario and the possibility of a move towards a plutonium based economy. By contrast, environmental groups acknowledged that new reactors would in themselves be inherently safer than existing designs (particularly the Magnox plants), but they placed greater emphasis on these other associated risks. They also argued that, while the risks may be remote, the impact of a major accident or terrorist attack could be colossal. The Government does in fact recognise the unique position of nuclear power in this respect, as it is the only generating technology to have its own police force.[129]

98. In considering these arguments, we found it difficult to assess the risks associated with nuclear and balance them against competing risks, such as those arising from climate change. At least in the West, the safety record of civil nuclear power plants has generally been good. There are many other chemical plants where accidents or terrorist attacks could potentially result in very serious impacts. This is demonstrated by the disasters at Seveso in Italy (where, in 1976, an explosion resulted in the release of dioxins), and at Bhopal in India (where a release of methyl isocyanate in 1984 resulted in nearly 4,000 deaths and 3,000 serious disabilities). Indeed, to the extent that security may be less onerous at such plants, the scope for terrorist attacks might even be greater. However, the potential scale of impacts resulting from a major accident or successful terrorist attack at a nuclear installation is of a different order of magnitude as the disaster at Chernobyl demonstrates. Calculations provided by the Oxford Research Group indicate that an attack on the high level waste tanks at Sellafield would dwarf the scale of the Chernobyl accident and could result in over half a million fatal cancers.[130] Moreover, the economic effects of such an event would be incalculable.

99. In this respect, the risks attached to nuclear power are indeed of a different order of magnitude to other forms of power generation. This is recognised in international law in the way that the insurance liability of the industry has been capped. The Paris and Vienna international nuclear conventions, negotiated originally in 1960 and 1963 lay down frameworks which shape most national nuclear liability laws and impose absolute but limited liability on nuclear site operators. They embody both financial and temporal limits for a catastrophic nuclear accident. In the UK, the financial compensation limit provided per site is currently £140 million and the temporal limit is 10 years—though the former is due to rise to Euros 700 million as a result of recent amendments to the conventions.[131]

100. We were also struck by the fact that the concerns expressed by environmental groups in relation to the proliferation of nuclear materials and technology worldwide appear to be shared at the highest level. The Oxford Research Group, for example, argued that the new Generation III nuclear reactors currently being proposed can use a mixed-oxide (MOX) nuclear fuel, which would allow countries and organisations easier access to plutonium. Moreover, Generation IV reactors (research on which is being conducted by an international collaboration of states, including the UK) would use mainly plutonium and would therefore constitute an even more serious threat. ORG cite a 2004 UN report which warned "we are approaching a point at which the erosion of the non-proliferation regime could become irreversible and result in a cascade of proliferation." While the UN report did not directly address threats arising from civil nuclear power, it did include the following general statement:

We recognize that nuclear energy, in the view of many, is an important source of power for civilian uses and may become even more crucial in the context of a worldwide effort to reduce dependency on fossil fuels and emissions of greenhouse gases. At the same time, the mounting tension between the goals of achieving a more effective non-proliferation regime and the right of all signatories of the Treaty on the Non-Proliferation of Nuclear Weapons to develop civilian nuclear industries needs to be addressed and defused. [132]

101. Such concerns are overtly demonstrated by Western attempts to constrain Iran's nuclear programme. Indeed, Professor Rogers argued persuasively that we may be on the threshold of a major escalation in the proliferation of nuclear technology. "Essentially, we are at a point where the nuclear weapons proliferation problem is possibly about to get much worse. That is something which is happening for political and strategic reasons. The problem is that in those circumstances simultaneously you are moving to a more heavily involved plutonium economy. Put that with the political and strategic problems, and you have a really serious issue over the next 10 to 20 years." He went on to emphasise that Western countries had no authority to persuade developing states to forego nuclear when they were seen to be saying 'Do as we say and not as we do.'[133] As Professor Rogers pointed out, the Director General of the International Atomic Energy Agency, Mohamed ElBaradei, is sufficiently concerned about these issues as to issue a seven point plan in an attempt to shore up the Non-Proliferation Treaty (NPT).[134] Yet the likelihood of substantive progress is minimal, particularly given the threats to the treaty not only from countries such as Iran (which is at least a member of the NPT), but also from India (which has refused to sign the NPT). Indeed, the recent agreement between the US and India to cooperate on nuclear development further undermines the NPT, and makes it is difficult to see how the US can argue with any consistency for upholding it in other cases.

102. In considering these complex issues, however, it is important to acknowledge that other generating technologies also have serious risks—albeit sometimes of a different type —associated with them. In particular, the exploitation of coal resources worldwide has resulted in countless thousands of deaths from mining and associated illnesses such as pneumoconiosis. Indeed, such impacts continue to this day in developing countries such as China and need to be balanced against the possible risks from exploiting nuclear power. Moreover, the development of new technologies—such as a shift to a hydrogen economy—can introduce other potential dangers, and we therefore welcome the review of safety issues which the HSE is now conducting. We trust it will assess the risks associated with all technologies in an objective and even-handed way.

103. The risk of a major accident at a nuclear power plant may be remote but the consequences can be huge. This is reflected in the need for governments to underwrite the industry against losses in excess of Euros 700 million. Moreover, the risks of terrorist attacks on nuclear installations and the risks associated with any further proliferation of nuclear power are serious.

Long-term waste disposal

104. The issue of nuclear waste is also an emotive and unique aspect of nuclear power, and one which is complicated by the historical legacy arising from the past 50 years of nuclear activity in the UK. The Nuclear Decommissioning Authority has now been established to manage the waste legacy, and the costs of doing so are currently estimated to be in the order of £60 billion. It was no part of our inquiry to investigate these issues, but rather to assess the prospects for dealing with waste from a new generation of nuclear power plants.

105. Currently, most spent fuel from nuclear power stations is reprocessed in Sellafield and temporarily stored there.[135] In 2000, the Royal Commission on Environmental Pollution recommended that no new nuclear power stations should be built until a solution to the problem of nuclear waste is found which commanded the confidence of both experts and the public.[136] The Government subsequently set up the Committee on Radio-active Waste Management (CoRWM) to investigate long-term storage options for high level and intermediate level waste and a final report is due in mid-2006.[137]

106. In their evidence to our inquiry, the nuclear industry argued that the impact of waste from a new series of nuclear new build to replace existing capacity would only increase the existing volume of all nuclear waste—including high level, intermediate, and low-level waste—by 15%, and that the costs of dealing with it would therefore be marginal.[138] The claimed increase in overall volume is correct and is based on CoRWM inventory data. However, further investigation of the CoRWM data reveals that the amount of high level waste would actually increase by 400%, thus necessitating some increase in the capacity of long-term storage solutions.[139] We discuss later in this report the financing of waste disposal from new nuclear build, but it is worth pointing out here that CoRWM's work is primarily directed to identifying the most appropriate long-term storage option for existing wastes rather than evaluating issues, including costs, associated with wastes from a new generation of power stations. CoRWM has itself stated that future decisions on new build should be subject to their own assessment process, including consideration of waste.[140]

107. Other countries face similar problems in dealing with waste disposal solutions. In France, in the late 1980s there was an abortive attempt to identify long-term disposal sites which resulted in local riots and the imposition of a 15 year moratorium pending further research. This report, like CoRWM's, is now due in mid-2006. Finland has identified a site and is going ahead with construction plans, even though the research reports evaluating the effectiveness of the proposed solution will not be completed until [2009]. These reports are of particular importance because there is no dry deep-storage anywhere in the country. In the US, the much-vaunted facility at Yucca Mountain is unlikely to be able to deal even with all the waste from existing power stations, and the industry is now proposing an additional site at Skull Valley. Moreover, the scale of the problem worldwide in the event of a 'nuclear growth' scenario is daunting as the MIT concluded in 2003 that a new long-term disposal site of the size of Yucca Mountain would need to be built every 3 years. These repositories would also need to last for thousands of years—a lifetime far longer than that of any nation which has hitherto existed on earth.

108. No country in the world has yet solved the problems of long-term disposal of high-level waste. The current work being conducted by CoRWM will not be sufficient to address the issue of waste associated with new nuclear build. In particular, a further study to identify the likely costs of the latter would be required in order to reduce investment risk.

89 Thomas, The Economics of Nuclear Power: Analysis of Recent Studies, July 2005.The paper can be found at:www.psiru.com. Back

90 Ev271, paragraph 43 and following table. Back

91 PIU Energy Review Working Paper, "The Economics of Nuclear Power", February 2002.See also National Audit Office, The sale of British Energy, HC 694, 1997-98. Back

92 eg Ev162. Back

93 For example, Steve Thomas suggests that the planning stage in Japan can take up to 20 years. Back

94 QQ110-111. Back

95 Thomas, The Economics of Nuclear Power, PSIRU, 2005. Back

96 Ev185. Back

97 Thomas, The Economics of Nuclear Power, PSIRU, 2005. Back

98 Ev269, paragraph 27. Back

99 TVO press notice dated 25 January 2006, at http://www.tvo.fi/757.htm The delays "are due to delays in detailed engineering and to the need to correct deviations in the manufacture of components." Back

100 Ev40. Back