How safe is that nuclear reactor?

Materials World magazine
7 May 2012
radiation symbol

Professor George Smith is Emeritus Professor of Materials at The University of Oxford, UK. He outlines his thoughts on the future of the nuclear industry and the problems that need addressing for energy security.

Turn the clock back 12 months and the world’s attention was firmly fixed on the risks associated with nuclear power, thanks to the disaster at Fukushima. Combinations of design and construction faults, external environmental impacts, and human factors principally caused the disasters at Fukushima and previously at Chernobyl. Meanwhile, issues associated with the performance of the materials used in the reactors have receded into the background. But with the evolving requirement to construct civil nuclear reactors with a safe operating lifetime of 60 years or more, materials issues need to be brought to the forefront once again. We must be confident that the structural materials used will retain their integrity. But we do not have 60 years to carry out the necessary experimental work. How best to proceed?

A key aspect of the answer lies in the development of an improved mechanistic understanding of the processes that lead to the degradation of materials properties, such as irradiation-induced embrittlement, long-term thermal ageing and environmentally assisted cracking. If these processes are understood at a sufficient level, then we can be more confident in predicting the future performance of the next generation of reactor systems.

But first we need to look in more detail at the global nuclear industry, and at the UK’s overall power generating requirements in the next 10–20 years:

The present global position on civil nuclear power

434 The number of operable civil nuclear reactors worldwide (including
104 in the USA, 58 in France, 51 in Japan, 33 in Russia, 21 in South
Korea, 20 in India, 18 in UK and 15 in China)

61 Reactors under construction (including 26 in China, 9 in Russia, 6 in India, and 5 in South Korea)

156 Further reactors planned (including 51 in China, 17 in India, 14 in
Russia, 7 in USA, 6 in South Korea, 6 in Poland and 4 in the UK)

343 More reactors proposed (including 120 in China, 40 in India, 30 in
Russia, 27 in USA, 16 in Saudi Arabia, 11 in Ukraine, and 9 in the UK)

14,750 Years of cumulative civil nuclear reactor operation to date (in 30 countries)

14% Of global electricity is produced from nuclear power (2010 figure).

(Data courtesy of the World Nuclear Association, January 2012,

In contrast to the highly adverse reaction in certain parts of western Europe (notably Germany) to Fukushima, the response in the rest of the world has been more cautious and considered. Efforts will be made worldwide to learn from the event and improve the robustness of reactor defences against natural or man-made disasters, but the pace of civil nuclear development is unlikely to slow significantly. The main impetus comes from outside western Europe, with China, India and Russia being the leading countries involved.

It is interesting to note the major nuclear incidents in the context of the overall use of nuclear reactors for power generation. Seven events stand out, not all of which led to the release of nuclear radiation:

1957 Windscale, UK – fire

1979 Three Mile Island, USA – near-meltdown

1986 Chernobyl, USSR – explosion

1999 Tokaimura Plant, Japan – criticality accident

2002 Davis-Besse Reactor, Ohio, USA – severe corrosion of pressure vessel

2004 Mihama, Japan – steam pipe failure leading to multiple fatalities

2011 Fukushima, Japan – multiple meltdowns following earthquake and tsunami

What is most striking is that there has been, on average, one major incident per 2,000 reactor operating years. In some industries, that would be regarded as an acceptable safety record. However, in view of the major long-term risks associated with the release of radiation, this level of performance within the nuclear industry is not acceptable to our society. It is impossible to quantify what level of risk is acceptable, but the industry needs to be aiming for at least one or two orders of magnitude improvement. This is a sobering thought. To achieve a failure rate of once per 20,000–200,000 years of reactor operation is an enormous challenge, particularly when economic demands are such that individual reactors are being required to remain in operation for longer than their initial design life, and are therefore at greater risk of long-term materials degradation.

Another feature of the accident data is the diverse causes of failures, including design flaws, materials degradation, engineering component failures, more complex system failures, operator errors, maintenance neglect, unforeseen behaviour of systems and external environmental disasters.

A wide-ranging total system approach to safety and risk assessment is needed, covering the whole lifecycle of the reactor, from initial design through construction to operation and maintenance, and eventual decommissioning. Passive safety systems have already been developed, but enhanced, multilevel systems of this kind may be needed for future plants to minimise the risk of human error, and allow for the possibility that a whole raft of active safety systems may be disabled by a single catastrophic event.

There is also a progressive change of emphasis in the evaluation of plant design. The very long projected operating lifetimes for new plants mean that the initial certification of plant safety is not an adequate guide to long-term outcomes. It is good practice to archive material at the construction stage, and to monitor materials condition through surveillance specimens periodically taken to measure mechanical properties. Such procedures have already allowed the operating lifetimes of UK reactors to be extended from 25 years to 40 years. But it is now necessary to go much further in this direction. Future plants need to be equipped with the engineering equivalent of a central nervous system so that if they are suffering from health problems of any kind, the plant operators can rapidly be made aware and can swiftly take remedial action. This requires developments in a range of technologies, from novel sensors and hardened electronics to complete nondestructive testing systems.

Reactor designs need to be evaluated not only on structural integrity grounds but also, increasingly, on maintenance criteria such as ‘get-at-ability’. Can remote inspection equipment be introduced into the heart of the reactor without excessive difficulty? And can the necessary repairs be carried out by the equivalent of keyhole surgery, rather than requiring long periods for dismantling and recommissioning? In an era when downtime costs may be up to £1,000,000 per day, value judgements about reactor designs will be increasingly dominated by such concerns.

It is also critical to have an adequate supply of fully trained manpower. Arguably this is the most important single factor in dealing with a disruptive event – and the recruitment and retention of key staff is already a major issue. With projected plant operating lifetimes far exceeding the working life of any single employee, techniques for the preservation and maintenance of accumulated knowledge and experience require as much attention as the maintenance of the plant itself.

The UK situation

The overall supply position for electricity production in the UK is extremely worrying. In 2007, dire warnings were being issued about the projected steep decline in baseload generating capacity, due to the need to decommission all but one of the UK’s ageing fleet of nuclear reactors by 2025, and the parallel need to remove from service a number of coal-fired plants that do not meet EU emission standards. This has led to independent predictions that the UK’s electricity generating capacity will fail to meet consumer demand within the next few years (Energy Materials Strategic Research Agenda, Materials UK Report 2007). Because of the unpredictable and intermittent nature of wind energy, it is unlikely that power from such a source will be sufficient to fill the generating gap. Experience from countries such as Denmark, where wind power has been exploited for many years, shows it is essential to provide around 90% back-up capacity from conventional sources. There is little sign the UK will be able to do this.

It is therefore vital that the UK should press ahead with plans to build a new generation of civil nuclear reactors. Put simply, the energy security of the country depends on it. However, there are serious challenges to overcome. The UK allowed its research and development investment in the nuclear area to decline to virtually negligible levels (see below) until recent efforts to revive it, for example at the Dalton Institute at The University of Manchester, Imperial College and the newly formed Bristol-Oxford Nuclear Research Centre.

As mentioned above, the supply of trained staff has plummeted (see above), and the age distribution of the remaining staff is towards older age groups. These resources will take even longer to rebuild than the research infrastructure. As a preliminary step, a knowledge capture exercise would be hugely valuable for preserving the vast store of experience and skill that was built up in the UK nuclear industry in the last 50 years, and which is in danger of being lost almost completely. But so far, no responsible UK body has shown the slightest interest in funding such a study – a tragic situation.

The previous Government’s sale of the last UK major nuclear manufacturing capabilities has made things worse, and the withdrawal by the present Government of the offer of a loan to Sheffield Forgemasters was a further blow to the industry. A recent House of Lords Select Committee Report, published in November 2011, provided a devastating critique of UK nuclear research and development policy failures over many years.

The only viable way ahead for the nuclear industry in the UK seems to be to form partnerships and alliances with other international participants in the field. The proposed agreement between Britain and France for collaboration in nuclear engineering, which was announced in mid-February 2012, is a potential lifesaver. Hopefully this will prove to be a turning point, safeguarding jobs and research activities in the UK, and allowing our remaining expertise in this field to be exploited profitably, and to the full.

A particular area of expertise developed in the UK is in the operation of advanced gas-cooled reactors (AGRs). These reactors are almost unique to the UK, and operate at much higher temperatures than the more conventional pressurised water reactors (PWRs) or boiling water reactors (BWRs). High temperature operation is likely to be a characteristic of the next generation of fi ssion plants (the socalled Generation IV reactors), so the accumulated UK experience in this area could prove especially valuable. Once again, however, politics has intervened. The last Government misguidedly withdrew the UK from the international Generation IV Design Consortium, meaning the UK is a spectator rather than a player in this key area of development. Hopefully, the new partnership agreement with France will be accompanied by a policy decision to rejoin this key research and development effort. Such a decision would also benefi t the development of a nuclear fusion power generating plant, which will have to operate under more extreme conditions than the proposed Generation IV fission reactors.

Materials issues

One of the key strengths of the UK’s research and development efforts in the nuclear field has been in the area of structural integrity. To ensure maximum safety for the next generation of PWRs, attention needs to be focussed on the following key issues:

  • embrittlement of pressure vessels and cooling pipes due to irradiation and thermal ageing
  • environmentally assisted cracking of pipes and welds
  • corrosion and erosion of cooling systems
  • creep-fatigue-irradiation interactions
  • thermal fatigue of large structures
  • ageing behaviour of joints, especially between dissimilar materials
  • long-term degradation of surface coatings and bearing surfaces
  • very long term degradation of waste storage container materials

Here, I emphasise two issues. The first is the opportunity presented by the current round of decommissioning of older nuclear reactors in the UK. This provides a unique chance to harvest materials that have been exposed to reactor operating conditions for extended periods of time. Such material is priceless. Forensic examination of this ‘real-life’ material can potentially give a much-needed reality check on the laboratory materials development programmes for future plants. Yet again, however, vital policy decisions are lacking. It has been estimated that systematic harvesting would add about 1% to the overall cost of decommissioning a worn-out reactor. But the UK’s Nuclear Decommissioning Agency has no remit to undertake such retrieval work, and no budget for it, so this unique opportunity seems likely to be missed unless urgent action is taken at the highest level.

The second issue is the urgent need to speed up the re-building of the UK’s infrastructure for nuclear materials research. Access by university researchers to facilities for handling radioactive materials remains extremely limited. This is a major handicap in training the next generation of nuclear engineers and materials experts. There are no good-quality neutron irradiation facilities available anywhere in the UK, and even support for existing research infrastructure is under threat. For instance, the 3D atom probe technique is vital for the understanding thermal ageing and embrittlement of reactor materials at the atomic level, but a request for longer-term support for the UK’s only facility of this kind, at the University of Oxford, has been under consideration by the Engineering and Physical Sciences Research Council for more than two years, with no sign yet of a final outcome. Such indecision threatens the whole future of nuclear materials research in the UK.

So, how safe is that nuclear reactor? My answer is – as safe as those running it. It is not sufficient for the UK, or any country, simply to go out, purchase the equipment that is needed and switch it on. The on-going safety and security of such systems requires the very highest level of skills, especially in materials science and engineering. We have to rebuild the physical infrastructure for nuclear materials research within the UK, but even more importantly we have to rebuild the human infrastructure as a matter of urgency.

Background reading

Energy Materials Reports: (1) Strategic Research Agenda, (2) Fossil-fuelled Power Generation, (3) Nuclear Energy Materials, (4) Alternative Energy Technologies, (5) Energy Transmission, Distribution and Storage (Materials UK Energy Review 2007).

The Mapping of Materials Supply Chains in the UK’s Power Generation Sector, S.A. Court, (Materials UK Energy Review, 2008).

Materials R & D for Nuclear Applications: The UK’s Emerging Opportunities, A.H. Sherry, C.A. English and P.E.J. Flewitt, (Materials UK Energy Review, 2010).

Japanese Earthquake and Tsunami: Implications for the UK Nuclear Industry, M. Weightman.

Nuclear Research and Development Capabilities. Report of the House of Lords Select Committee on Science and Technology, (22nd November 2011) [HL Paper 221] (

Generic Assessment of Candidate Nuclear Power Plant Designs: Interim Statements of Design Acceptability for the UK EPR and AP 1000 Designs (and supporting documents) (Environment Agency 14th December 2011).

Nuclear Reactor Materials at the Atomic Scale, E.A. Marquis, J.M. Hyde, D.W. Saxey, S. Lozano-Perez, V. de Castro, D. Hudson, C.A. Williams, S. Humphry-Baker and G.D.W. Smith, Materials Today, vol. 12 no, 11, November 2008, pp. 30 – 37 (and references therein).

A Comparison of the Structure of Solute Clusters Formed During Thermal Ageing and Irradiation, J.M. Hyde, G. Sha, E.A. Marquis, A. Morley, K.B. Wilford and T.J. Williams, Ultramicroscopy vol. 111 no.6, pp. 664-671 (2011) (and references therein).

Further information

Professor George Smith:

The author would like to thank his Oxford colleagues, along with
Professors Peter Flewitt, Andrew Sherry and Colin English, Dr Jonathan
Hyde, and Dr Tim Williams for their contributions