Reaction to change

With nuclear power of the future expected to embrace a far wider range of applications than today, researchers are investigating the technology necessary to run new-style reactors. Stuart Nathan reports.

While the nuclear industry is preparing for new-build to meet a looming energy gap, researchers have their eyes on the future. Although the next-generation power stations will be far more advanced than their predecessors, they will be based on the same reactor technology.

In the coming decades there are likely to be huge changes in how we think of energy — not just electricity but all the ways that we generate and use heat. Nuclear power could have a far wider range of applications than its current uses for baseline electricity generation and as a power plant in naval vessels. It could also form the basis for the shift from fossil fuel-based to hydrogen-based transport.

For this, new types of nuclear reactors will be needed. The existing generation of commercial reactors, mostly based around pressurised water cooling technology, are fine for power generation. The likely candidates for the new build of nuclear power now mooted have similar reactor systems to their predecessors, with refinements to reduce cost and improved safety features. But they are optimised to work non-stop at full capacity and have no function other than to raise the temperature of water high enough to run a steam turbine. Future reactors will have to be far more versatile.

Reactors are classed in terms of generations. Generation I comprised the first experimental reactors. Generation II were the first industrial-scale commercial reactors, including the early Magnox stations in the UK. Generation III are the current commercial reactors, generally using pressurised water reactor (PWR) technologies. The newest plants are known as Generation III+, and include systems that prevent meltdown in the case of a serious incident without intervention from the plant’s operators.

The next wave of plants, known as Generation IV (G4), are in most cases still on the drawing board, but are the subject of intense research around the world. They are probably about 30 years away from commercial exploitation and are being designed according to stringent criteria which, nuclear researchers believe, reflect the changing conditions for energy generation.

They are to be more economically competitive than their predecessors, with better safety features including refinements of passive safety; they should make more efficient use of uranium resources and produce as little waste as possible; their vulnerability to physical attacks will be minimised; and they will not contribute to nuclear proliferation.

At least, those are the goals of the Generation IV International Forum(GIF), a 13-member international organisation that oversees development work on these advanced reactors.

Established in 2001 with nine members — France, Argentina, Brazil, Canada, Japan, South Korea, South Africa, the US and the UK — GIF co-ordinates collaboration on the design and development of reactors, although each country funds its own R&D, decides which systems it will work on, and retains shares in the intellectual property resulting from the research.

Because of this, it is difficult to put a value on the funding for G4 research, although as it is still mainly conceptual in nature, the amounts are relatively modest; in 2006, for example, the US spent $39m on G4 research. Switzerland, China, Russia and the Euratom nations have joined GIF since its inception.

One of GIF’s first jobs was to select which designs its members would devote their time to. The 100-strong team of engineers from the original nine members whittled the 130 designs submitted from nuclear researchers around the world down to six.

Three are ‘fast reactors’, which use one of the products of the fission of uranium-235, an energetic or fast neutron, to transmute other elements into isotopes that can undergo fission. They are also known as breeder reactors, as they ‘breed’ their own fuel; this is also known as a closed fuel cycle, because no new fuel needs to be mined once the reactor starts operation.

They could also have another important role, as their ability to induce nuclear reactions means that they can transmute hazardous radioactive waste into elements with lower radioactivity, a process known as ‘burning up’. The other three types, known as thermal reactors, need to be refuelled in the same way as today’s reactors.

Two of the six reactor types are revivals of older technologies which, for a variety of reasons, had been neglected or mothballed. The sodium-cooled fast reactor (SFR) submerges its fuel elements in a bath of liquid sodium, while a separated sodium circuit carries away the energy generated by fission to heat water.

The very high temperature reactor (VHTR), meanwhile, uses fuel in a ceramic form and generates temperatures around 900°C, which increases its efficiency and allows it to be used for a multitude of non-electricity generating uses.

The four other reactor types include two more fast reactors — the gas-cooled fast reactor (GFR), similar to the VHTR but configured to generate new fissile materials, and the lead-cooled fast reactor (LFR) which uses molten lead or a lead/bismuth mixture, both of which are less reactive then sodium, as a coolant.

There are also two thermal reactors: the supercritical water reactor (SCWR), which does not need steam generators and has smaller turbines than other designs, and could therefore be a more economic option; and the molten salt reactor (MSR), whose fuel, a mixture of liquid fluorides of uranium and plutonium, is dispersed in the coolant, the molten salts of light metals such as lithium and sodium. This system requires less fissile material on-site than other designs, incorporates systems for processing wastes within the facility, and does away with the need for fuel elements.

The design for SFR is further developed than other G4 technologies because of previous R&D

While all of these are receiving attention from the GIF partners, development of the SFR and the VHTR are well ahead of the other four, since the current research picks up the threads of previous developments.

SFR technology dates back to the 1950s; the reactors at Dounreay in Scotland used earlier forms of this technology and other countries also operated sodium reactor programmes; these were mostly abandoned in the mid-1980s, when the end of the Cold War reduced the price of uranium and the reactors became uneconomic.

VHTR originated in Germany, where there was an active programme until the early 1990s when pressure from Green Party MPs led to the country abandoning nuclear power.

‘The experience with these types of reactors is very valuable, and in a sense that reduces the research and development needed for the completion of what we’d offer in G4,’ said Ralph Bennett, technical director of GIF and the director of international and regional partnerships at the Idaho National Laboratory (INL), one of the leading centres for VHTR research.

Despite previous research, developing G4 versions of these reactors is no easy task.

‘In the case of the SFR, for example, it’s not enough to say that we have a sodium reactor which has a different variety of fuel, or that we’ve demonstrated a bit of recycling of just the plutonium that’s in the reactor.

‘What we want in G4 is the ability to address the reduction of waste, and that takes a lot more time; for example, we need to qualify fuels of all sorts of compositions, including the exotic minor actinide metals and other materials.’

The major stumbling block for the original sodium fast reactors was economics, however; their electricity was simply too expensive. ‘A very central feature of the research being done is to find breakthroughs in the economics,’ said Bennett.

Research is also needed in materials science, as the reactor casing and internals have to withstand the bombardment of high-energy neutrons; unlike thermal reactors, fast reactors contain no moderator material to slow the neutrons down.

Sodium reactors are intended for power generation, as well as the treatment of nuclear waste. But as the experience the UK gained at Dounreay showed, they are a difficult sell: they turn non-fissile material into fissile, which can, in certain types of reactor, be used to make weapons-grade plutonium, although one of the goals of G4 development is to ensure that the technology developed is unattractive, and the least desirable route, for this application.

Moreover, the use of sodium carries with it inherent risks, and the problems of the past, including radioactivity leaks and the difficulty of decommissioning old reactors, has tended to turn public opinion against this technology.

Colin Bayliss, the operations director of UKAEA, is mindful of this. ‘We had our reactor [the Prototype Fast Reactor, PFR, at Dounreay] working 24 hours a day, generating about 250MW on to the grid, and they had fully closed the cycle for the core fuel, although not for the fuel it was breeding,’ he said.

‘But I gave evidence to a House of Lords select committee about seven years ago, where I said that I didn’t think fast reactors would come again for about 40 years.’

There have been great advances in materials science since the closure of Dounreay, he said, but problems remain to be solved.

The trigger for the renewed interest is an anticipated scarcity of fissile uranium, he said: current estimates say there is enough for about 70 years’ operation of the projected reactor fleet around the world.

‘And that fits in well for 40 years time, when we might be thinking of production models for future fast reactors. But it comes down to public opinion almost more than technology. People must treat the public acceptance issue in a serious, systematic way.’

Public acceptance for VHTR could also be a problem. While opinion is slowly coming back around to nuclear power for electricity generation, the main application for VHTRs is their ability to generate heat.

‘They will open up a new market for industrial process heat applications, including the use of this heat to make synthetic fuels, especially hydrogen,’ said Dominique Hittner, chair of the steering group for the EU’s VHTR project, RAPHAEL. ‘It’s also good for cogeneration, so you’ll be producing some heat and some electricity; cogeneration plants are common in industry, usually gas-fired today, and these represent another version.’

VHTRs are proposed to be fairly small reactors, with a thermal output of 400-600MW, less than half the output of a GIII+ reactor. Two types of reactor have been proposed. The smaller of these is the pebble-bed modular reactor, which is being developed in South Africa and is nearing a prototype stage. This encapsulates its fuel within tennis ball-sized graphite spheres, which contain thousands of ‘micro-pellets’ of uranium-235 surrounded by a shell of silicon carbide. The reactor is cooled by helium gas.

Pebble-bed systems will probably be the first G4 technologies to reach the prototype stage

The other type of reactor is known as a prismatic block, where hexagonal blocks of graphite contain fuel rods made up from a similar type of fuel configuration as the pebble bed, again cooled by helium. These types of reactor have inherent passive safety because, as the fuel heats up, a quirk of the uranium atoms in the fuel means fast neutrons, which can cause runaway chain reactions if they strike fissile atoms, are more likely to be safely absorbed by non-fissile atoms. At around 1600°C — well below the melting point of the fuel — the nuclear reactions stop.

One reason for the interest in VHTRs is that they can be used to make hydrogen using a thermal process. ‘You can make hydrogen by electrolysis, and any nuclear plant could provide that,’ Hittner said. ‘But using heat to make it could be more efficient.’

In the early stages of RAPHAEL, he said, the target outlet temperature for the coolant helium is 700°C-750°C; this could be high enough to provide the heat for producing hydrogen via a membrane methane steam-reforming process. ‘We are working with companies that are developing such a process which will work around 600°C,’ he added.

But boost the temperature higher, above 900°C, and a process based on a reaction between water, sulphur and iodine might become practical. This is a complex process and is not now possible at an industrial scale, but could be an efficient way to make large volumes of hydrogen without generating any carbon emissions.

Hittner is sceptical that such high temperatures will be needed. ‘We would have to do work on the materials and fuels that could tolerate the temperatures, if the market needed them,’ he said. ‘But there’s a general tendency to decrease the temperatures of processes if at all possible, to reduce energy consumption.’

There are many technical barriers for VHTRs. Hittner said as they will not operate in a grid, they need to have a far more variable power output that electricity-generating reactors. ‘Operators will want to have the power when they need it,’ he said. ‘So you might need at least two reactors on the site, with one as back-up.

‘But the advantage of high- temperature reactors is that they’re small; by adding as many units as you need, you gain flexibility for the total power. You’d run enough reactors to meet your requirement, and bring more on-line when more heat is needed.’

For the VHTR project, he added, the end-point of the research is not just the reactor, it is the integration of the reactor with all the other processes it is intended to power.

Prof Robin Grimes, Imperial College’s head of materials physics, pointed to other problems with VHTRs. ‘They have no intermediate heat exchanger,’ he said, explaining that in most reactors, this is a loop that extracts heat from the reactor coolant then passes that on to water, rather than using the reactor heat directly to raise steam.

‘The reason for that is that the coolant is helium, which doesn’t become radioactive. That’s fine, but we know that all kinds of elements can get created inside reactors and spall off the sides, and then the helium might act as a carrier gas.’

There are also issues with the way the waste would be recovered from the fuel pellets and treated. ‘A reactor without a clearly-defined waste path doesn’t have a chance now,’ he said.

Ralph Bennett insists waste issues are paramount for GIF. ‘We are concentrating on techniques and technologies that mitigate the unavoidable aspects of these reactors,’ he said. These include the generation of waste, and problems such as the difficulty of maintaining and inspecting the interior of SCR reactors, which are filled with hot, opaque liquid metals. ‘We have a senior advisory panel of a dozen or more high level executives from major utilities, including Areva, EDF and others, and they are very keen to stimulate research into all corners of the lifecycle of the plant, including decommissioning, supply and treatment of fuel and qualification of components.’

Bennett thinks the first G4 reactors are some time off. ‘2020 might be the earliest for the first VHTR or SFR reactors; the others, maybe around 2030. There might be pebble-bed prototypes up and running and generating electricity before 2020, and that will give a lot of fresh understanding, but if it doesn’t make hydrogen, and if you haven’t tackled the problem of developing the whole system, integrating it with the reactor and getting the whole thing licensed, then you haven’t crossed the finishing line for G4.’