Core decision

With the UK poised to embark on a new era of nuclear energy, what technologies will be used in the next generation of reactors? Stuart Nathan examines the options.

Britain’s fleet of nuclear reactors is close to sailing into the sunset. Of the 19 nuclear-powered power stations dotted around the coastline, 11 are no longer in operation and one more will cease generating in 2010. All are based on outmoded technology except Sizewell B in Suffolk — the only one that will continue operating beyond 2020. With the recent Energy White Paper coming down on the side of nuclear power, it seems certain the fleet will be replaced.

But what with? Current nuclear technology is far ahead even of Sizewell B and new reactors are being developed along strikingly different paths. The parties that will decide what sort of reactors will help keep the UK’s lights on — regulatory bodies, power generators, plant contractors, politicians, NGOs and communities — have some tricky choices to make.
Nuclear reactor technologies are described in generations. The earliest, Generation I (GI), are the Magnox reactors, such as the UK’s Calder Hall, the world’s first power-generating reactor, dating back to the 1950s.

Generation II reactors make up most of the world’s reactor fleet and include the advanced gas-cooled reactors (AGRs) used in the UK and the pressurised water reactors (PWRs) in the US. These have more standardised designs than the GI reactors, which were often experimental, bespoke designs. They are also much smaller.

A typical Magnox reactor is 9m high and 17m in diameter, while an AGR is about half the diameter and the PWR, derived from the power plant of a nuclear submarine, is 4m high and 4m in diameter. Sizewell B, the UK’s most modern plant, is a GII PWR, the only one of its type in the country.

The GI and GII plants operate in the same way — the heat generated from nuclear fission reactions in the fuel, which is based on uranium and contained inside the reactor, is carried away by a coolant, which is carbon dioxide (in Magnox and AGRs) or water (in PWRs). This heat is then used to generate steam from water, which is kept apart from any radioactivity. The steam spins turbines to generate electricity just as a steam turbine in a fossil-fuel powered generator.

Refinements in the design of PWRs has led to the development of Generation III reactors. Many of the refinements improve the safety of the plant, including multiple-redundancy systems to keep coolant circulating around the plant (coolant failure led to the accident at Three Mile Island in 1979 and the catastrophe at Chernobyl in 1986). Others improve the thermal efficiency of the plant in converting heat into electricity.

Further design improvements have given rise to a new generation of plants — the Generation III+. Still at the development stage, these have the same layout as a PWR but incorporate safety features that work differently to the multiple-redundant GIII systems.

The main thing that distinguishes GIII from GIII+ is a feature known as passive safety. In all reactor designs until now, the operation of the plant centres around controlling the nuclear fission reaction in the core and keeping the reactor safe: if the plant is left alone, a chain reaction will occur, which will go out of control and heat the core to the point where the fuel melts. Constant intervention is needed to stop this happening.

The concept of passive safety is to reverse this: to require constant intervention to keep the reaction going. If the plant is left alone, the reaction will stop and the core temperature will build to a peak level, below the melting point of the fuel, then begin to fall.

GIII+ designs are also simpler than GIII, require less material to build and have fewer systems.

Even more advanced are the conceptual designs known as Generation IV, which adopt different approaches to fuelling, controlling and cooling the reactor.
The UK’s new reactors are likely to be GIIIs, with some features that are claimed to liken them to the GIII+. Unlike the GI and GII reactors, they will not be designed in this country.

‘We had the capabilities and we proved them, but we lost them,’ said Robin Grimes, professor of materials physics at Imperial College, who is at the forefront of training a new generation of nuclear engineers.


Sizewell B is the only existing nuclear-powered station that will still be functioning beyond 2020. The two front-runners for the replacements are the AP1000 (above left) and the EPR (above right)

According to Grimes there are three main contenders, two from the US and one from Europe. In the US corner are the General Electric ESBWR (Economic Simplified Boiling Water Reactor) and the Westinghouse AP1000; from Europe is the Areva EPR (European Pressurised Reactor).

The EPR has an advantage over the others in that it is nearest to production. The first is now being built in Finland, at the Olkiluoto site; another is planned for Flamanville in France.
The designers of the EPR — EDF, Siemens/KWU, Framatome and Nuclear Power International — combined the best design aspects of the most recent French and German fleets.

Areva, a French company that is the only nuclear operator active in all parts of the sector, from mining uranium to dismantling reactors, claims EPR is a GIII+ design, but Grimes disagrees.
‘It’s a bog-standard PWR, very much of the Sizewell B variety. There is no real passive safety in the EPR, although it does have enhanced safety features.’

Chief among these features is a core catcher, a pit below the reactor made from a highly refractory (heat-absorbing) material where, in the event of a meltdown, the molten core would spread out and be cooled down by a reserve supply of coolant water.

It also has four redundant systems for safety injection, component cooling, back-up electrical power and steam generator emergency feedwater. These are housed in separate buildings, two of which will be in bunkers able to survive aircraft crashes.

But Grimes is dismissive of the core catcher concept. ‘It’s completely ridiculously over-engineered, pretty much unnecessary, and costs a lot of money,’ he said. ‘Although there’s no question that in the event of a very serious accident like a large loss of coolant, the EPR would provide an even safer outcome than older reactors — but at a significant monetary cost.’

Despite that, EPR is a strong contender, and Areva has applied to the Health and Safety Executive, whose Nuclear Inspectorate has to approve nuclear power station technologies, for pre-licensing of the EPR. Among the power generators endorsing the application is British Energy, which operates the UK’s nuclear fleet.

The other factor in the EPR’s favour is its lack of innovative features. Every aspect of the system is well understood and in operation, according to Paul Howarth, director of research at the Dalton Nuclear Institute at Manchester University. ‘There’s really nothing there which would make the regulators uncomfortable,’ he said.

This is not true of another of the front-runners, the Westinghouse AP1000. ‘Back in the 1980s, the Americans went right back to the drawing-board, trying to make the reactors simpler,’ Howarth explained.

‘The driving force was to get the capital costs down. The reactors were getting so big and so complicated, with redundant safety system upon redundant safety system, and they just weren’t economic any more.

‘Westinghouse developed a system that removed most of these redundancies, and relied instead on natural processes that can’t fail — gravity, convection, and natural circulation — to simplify the technology. That effectively has trailblazed some of the features as to what passive safety actually is.’

The reactor uses components that are tried and tested on Westinghouse PWRs — such as the fuelling system, the pressure vessel and the coolant pump — then adds on passive safety systems.

For example, the safety systems to cool the reactor are powered entirely by gravity. ‘There’s a massive tank of water on top and water flows downhill,’ said Grimes.

The systems use no pumps, fans or rotating machinery. All the valves in the safety system require power to keep them closed and, in the event of a loss of power, will lock open to allow coolant to circulate.

‘The AP1000 incorporates passive systems where, if you had a loss of coolant accident, there would be a period of time over which the fuel would maintain its integrity,’ said Grimes. ‘But if you don’t do anything, eventually you will get core melt. It’s quite a large period of time; depending on the type of accident it’s many, many hours or many, many days. But there’s no doubt in my mind that it would work.’

Another attractive feature of the AP1000, said Grimes, is that it uses a third of the land space taken up by a conventional PWR. According to Westinghouse the nuclear ‘island’ of an AP1000 station is small: it has half the pumps, 35 per cent fewer valves, 80 per cent less piping, 85 per cent less cabling and 45 per cent less seismic building volume than a conventional PWR.

Even with the turbine hall, which is exactly the same size as any other power station of its generating capacity, this still leads to a considerably smaller footprint, less materials, and therefore quicker building time and lower cost.

‘The promotion is that it’s even safer [than previous PWRs], smaller, cheaper and simpler,’ said Grimes. ‘I think simpler is very important, and I think you could even argue on the basis of it having a smaller carbon footprint to construct.’

But these advantages are all theoretical, because although the AP1000 is approved by the US regulator and has been submitted for approval in the UK, it has not been built yet.
‘Some of these safety features are novel and the regulator will have to assess them to see whether it’s comfortable with giving approval,’ said Howarth.

‘It depends how much credit the regulator gives to US licensing. Tests have been done: Westinghouse has carried out tests at a quarter scale, but there are definitely issues to be addressed.’

The third contender, from engineering giant GE, takes yet another approach to power generation technology. The ESBWR, unlike other reactors, does not have a separate circuit for the water and steam that spin the turbines — the reactor itself boils the coolant to raise steam, which then flows straight into the turbines, before being condensed and returned to the reactor in a closed circuit.

‘It’s cheaper because it doesn’t have a secondary heat exchanger, but that balances out because the steam in the turbines is activated, so you have to have that within containment as well as the reactor,’ said Grimes.

The ESBWR also incorporates some passive safety, such as isolation condensers, which take steam from within the process, condense it, and return it to the reactor vessel. A gravity-driven cooling system lets water into the vessel if it detects low coolant levels, and a piping structure below the reactor will catch, divide up and cool down molten coolant in the event of a meltdown.

Grimes, however, is not a fan. ‘They have all sorts of issues, and I don’t think they represent such a good choice, but GE might have the clout and ability to push it through.’ Howarth, too, believes Westinghouse has the edge in terms of technical novelty. ‘ESBWR has a few novel features, but if you were saying which is truly innovative, it’s the AP1000.’

With the timespan available to build new plants, Grimes believes that the most likely options are the AP1000 and the EPR. Which are built depends largely on the consortia that will build and operate them. ‘EDF is never going to build anything but an EPR, is it?’ said Grimes.

A further consideration with EPR is that, with two plants already under construction and more sure to be built in France, any orders from the UK will need to be relatively quick, or they would be so far down the chain that supply delays would be inevitable.

What should not be in doubt, Grimes stressed, is the UK’s ability to run PWR plants such as these.

‘Sizewell is extraordinarily well-run; one of the best in the world, in fact. It shows that British Energy, if it’s given the tools, can do a very, very good job.’

SIDEBAR: Waiting in the wings

While the next generation of reactors will be based on existing technologies, the one after that is likely to look very different. True GIII+ technologies are some 10-15 years off, said Imperial College’s Prof Robin Grimes, with Generation IV designs — incorporating more passive safety systems, producing less waste and operating more efficiently — now under development.

In South Africa, the pebble bed modular reactor (PBMR), a GIII+ design, is in a test phase. Smaller than PWRs, the PBWR does not have fuel rods like a conventional reactor. Instead, the fuel consists of many particles of uranium oxide coated in pyrolytic carbon and silicon carbide, encased within a sphere of graphite about 60mm in diameter. When fully loaded, the reactor core would contain 450,000 spheres.

The reactor is helium cooled, but in the event of a loss of coolant, there would be no meltdown. The surface area of the fuel spheres is so large compared with the volume of the reactor they would lose heat faster than the fission inside the spheres would generate it. The maximum possible temperature inside the core is 1600°C, its developers claim, well below the temperature that would damage the fuel.

‘My experience of the PBMR is that it’s a great idea but there are all sorts of things they’ve got to pay attention to,’ said Grimes. ‘There’s no clear way they are going to reprocess or deal with the waste at the end, and in a modern reactor, you should design for the whole life-cycle.’

Generation IV reactors also take a different view on reactor design. In 2002, a consortium of 100 experts from nine countries — Argentina, Brazil, Canada, France, Japan, South Korea, South Africa, the UK and the US — along with the European Union, came together as the Generation IV International Consortium (GIF) and designated six reactor designs for research and development.

These are the molten salt reactor, which uses a liquid mixture of sodium, zirconium and uranium fluorides as a coolant; the lead-cooled fast reactor; the sodium-cooled fast reactor; the supercritical water-cooled reactor; the gas-cooled fast reactor and the very high temperature reactor.

Grimes believes the very high temperature reactor (VHTR) is particularly interesting, as its high operating temperature of just under 1000°C could be used to thermally crack water to produce hydrogen.

‘Also, graphite technology is key to this design, and what does the UK have in abundance? Graphite experience, more than anyone else. All our AGR reactors are graphite-moderated.’
However, he added, the UK has not ratified the GIF agreement. ‘It would cost such a small amount of money, and with that degree of expertise waiting in the wings, it’s extremely disappointing.’