Thorium: big hopes, small steps - nuclear fuel

Materials World magazine
1 Sep 2013

Boosted by the recent establishment of the Weinberg Foundation, thorium nuclear power is enjoying an increasingly high profile as a green nuclear energy panacea. The reality, as Tim Probert discovers, is not nearly so simple.

Click here to read a PDF (1.7mb) of this article with full diagrams. (Low res version PDF 400k)

The profile of thorium as a potential source of nuclear power has rarely been higher. An increasing number of gushing articles are appearing in newspapers, magazines and on websites extolling its virtues as a green source of nuclear power. Believe what you read and it’s easy to see how the general public could assume thorium is something of an energy magic bullet, offering all the benefits of atomic power without downsides such as potential core meltdowns and copious amounts of radioactive waste.

The truth is, of course, far more complex. While thorium offers highly attractive potential advantages, there are also several considerable barriers to the development of thorium-based nuclear reactors. The main obstacle facing the development of thorium as a nuclear fuel is uranium. Virtually the entire nuclear energy industry has been optimised around the uranium and, to a lesser extent, plutonium fuel cycles.

Thorium-232 (T232) is a fertile, not fissile, isotope. T232 has to be irradiated in a reactor to produce its derivative fissile isotope uranium-233 (U233). To begin its nuclear weapons programmes in the 1940s, the USA decided to produce enriched uranium in U235, the only naturally occurring fissile isotope in nature, as well as plutonium-239 (Pu239), from neutron irradiation of fertile natural uranium (U238) in a reactor. The USA recognised but declined the third option to produce U233 from irradiation of Th232 in U235-enriched reactors for two main reasons.

First, it was known that U233 was a hard gamma emitter that would be difficult to reprocess, to machine and to handle from a personnel protection standpoint. Second, and more importantly, it was demonstrated that the 0.7% level of fissile U235 in natural uranium was sufficient to be the basis for a nuclear reactor system, with the 99.3% U238 content in natural uranium hence being partially converted to produce around 0.8% of the new fissile element Pu239 – unknown in nature – which could then be chemically separated through reprocessing.

Had the USA gone down the thorium route, there would have been a significant weapons programme delay in producing sufficient quantities of the required enriched uranium driver. Uranium, therefore, got a head start for civil nuclear power from which thorium has never recovered. The vast majority of the 434 operational nuclear reactors worldwide are light water reactors (LWR), developed as scaled-up versions of nuclear submarine reactors.

Although the USA’s research at the Oak Ridge National Laboratory (ORNL) in Tennessee, which produced plutonium for the Manhattan Project, ultimately gave birth to the prevailing LWR uraniumfuelled nuclear power industry, it has also left a lasting legacy that offers the promise of a radical alternative. The molten salt reactor experiment (MSRE) is something of a cult project. Led by Alvin Weinberg, research director of Oak Ridge National Laboratory, which produced plutonium for the Manhattan Project, MSRE was a 7.4MW demonstration reactor using U235 and later U233 as the main fissile driver. While not strictly a thoriumbased project, MSRE proved the potential for molten salt reactors (MSRs). Unlike conventional nuclear reactors that use solid fuel, usually rods or pellets, MSRs use a mixture of fluoride salts in a molten state. The salt mixture includes fissile material – isotopes of uranium or plutonium, together with fertile material such as T232 or U238.

Perhaps the most promising thorium-based MSR is the liquid fluoride thorium reactor (LFTR). In this design, the molten salt also serves as the primary coolant, carrying heat away from the reactor, and delivering it to a secondary cooling circuit and ultimately to the steam turbines that generate electricity. One of the main advantages is the reactor and its cooling circuits operate at near atmospheric pressure, reducing the chance of any explosion. As fuel in a LFTR is already in liquid form, it cannot melt down, as solid fuel rods can in a LWR.

Furthermore, the LFTR’s large negative temperature coefficient means that regulation of the reactor’s temperature is passive, so there is no need for control rods. The molten salt expands as a result of the heat generated by fission, which in turn slows the rate of fission. The reduction in fission then cools the salt, which leads to an increase in the rate of fission. In other words, as the reactor temperature rises, the reactivity decreases. The reactor thus automatically reduces its activity if it overheats. In the event of a reactor overheating, the fuel and salt drain into a holding tank, where the fuel spreads out enough for the reactions to stop. The salt then cools and solidifies, encapsulating the radioactive materials.

Another major advantage over uranium is fuel efficiency. Due to the degrading effect of neutron bombardment on solid fuel-rod metal cladding in LWRs, between only 2–4% of the energy contained within can be used before the rods have to be removed. An LFTR, however, would continuously recirculate nuclear fuel, greatly improving efficiency and radically diminishing the produced volume of nuclear waste and proliferation risk.

On paper at least, LFTRs are highly attractive and it is easy to see why they have a growing number of proponents. In practice, however, there are major drawbacks yet to be overcome. While Weinberg’s work at ORNL’s relatively short-lived MSRE successfully demonstrated the potential for thorium MSRs, a number of serious problems were highlighted, particularly pertinent for power generation reactors with an operational lifetime of several decades.

Researchers found Hastelloy-N, a low chromium, nickel–molybdenum alloy developed by ORNL for the MSRE reactor vessel and pipework, cracked and corroded under intense radiation at high temperatures of up to 650°C. The development of materials capable of surviving exposure to neutron bombardment and highly corrosive fluoride salts at temperatures of 650– 800°C or more for several decades and is, therefore, essential if MSRs are ever to be a commercial proposition.

Alternatives to Hastelloy-N

There are several potential alternatives to using nickel-based alloys such as Hastelloy-N, but each has considerable drawbacks. The reactor vessels of existing LWRs are usually forged from steel or stainless steel, but the maximum operating temperature is 450–600°C – too low for the higher operating temperatures of MSRs.

Titanium alloys are one option. Titanium has a high strength-to-weight ratio, as well as good temperature, creep and corrosion resistance. Titanium in its liquid form, however, is highly reactive, making it necessary to process it in a vacuum, a highly problematic proposition. Other options include using tungsten and tungsten alloys. Pure tungsten has an extremely high melting point of 3,410°C as opposed to 1,725°C for titanium and 1,458°C for nickel, and has high creep resistance and neutron flux resistance. However, tungsten is somewhat brittle and prone to shock load failure. Perhaps most importantly, tungsten is extremely costly, trading at around £30/kg, compared to around £9/kg for nickel, £4/kg for titanium and just £0.08/kg for steel billets.

Ceramics such as silicon carbide are another option with high temperature resistance, but they are brittle and can unexpectedly fail and crack as a result of shock. It is also questionable whether silicon carbide would be sufficiently resistant to highly corrosive molten fluoride salts. Ceramics can also be expensive and difficult to form, which is problematic given the need to for an airtight seal to the graphite core.

For the foreseeable future, nickel alloys remain the least worst option. Research into MSR materials is limited, although French firm Albert & Duval, which specialises in advanced metallurgy for aerospace and energy industries, is participating in the EU’s Evaluation and Viability of Liquid Fuel Fast Reactor System (EVOL) programme. EVOL will see the design of an MSR from physical, chemical and material studies, including the development of improved Ni-W-Cr alloys. Work will focus on controlling the salt chemistry to minimise corrosion, essentially by controlling the redox potential of the melt. The core in the EVOL project will be a direct mixture of lithium fluoride and thorium fluoride, which means operating temperatures would be higher still than the Oak Ridge project. EVOL will therefore be exploring different compositions of nickel alloys with varying chromium content.

Research in the UK

While there is no direct research into MSR materials in the UK, it could piggyback on nuclear fusion research conducted at the Culham Centre for Fusion Energy in Oxfordshire, says Professor Paul Madden, Provost of Queen’s College, Oxford. He explains, ‘There is a lot of work being done in the UK on solid-state materials motivated by the nuclear fusion project. There is a large element of cross-fertilisation between the fusion-driven containment materials and MSR containment materials.

‘A fusion system would have a great deal of extremely fast neutrons and what you're looking for is a way of degrading that energy to heat. It may be that this is not very sensible, maybe you should be using those neutrons for activating materials.’

When the materials problem has been solved, there is still the problem of developing a resilient online chemical processing plant that continuously removes fission products during MSR operation (see MSR schematic). Unlike solid fuel LWRs, where the fuel is removed, treated, remanufactured and reinserted in the reactor, liquid fuel MSRs would need continuous removal of gases via an online processing system.

The MSRE at Oak Ridge was merely a 7.4MW unit. Considerable research is necessary to ensure the fluid dynamics of the online chemical processing plant can be scaled up to a circa 1GW power station.

UK thorium nuclear power research is limited, with total investment estimated at less than £1 million over the past five years. The University of Huddersfield is conducting research into an Accelerator Driven Subcritical Reactor (ADSR) based on thorium. In an ADSR, the critical flux of neutrons is created through a beam. A spallation source is bombarded with protons and this bashes off neutrons, creating a beam of neutrons. The neutron beam then hits the nuclear fuel, thus achieving criticality. Such a reactor would be inherently safe because it could be effectively turned off with a flick of a switch.

One of the many problems arising from the muchmaligned Sellafield MOX plant was the burning of weapons-grade plutonium created minor actinides such as americium, curium and neptunium, which can have half-lives of millions of years. Because of the type of neutron spectrum created using an ADSR, it could be used to burn up minor actinides.

The University of Huddersfield’s Professor Robert Cywinski says thorium will come into its own as a nuclear waste problem-solver. ‘We started off looking at thorium as a power generation solution, but we see the accelerator approach as an ideal way to dispose of our legacy waste. If you use a conventional fission reactor, you don’t have the right energy neutrons for fissioning minor actinides. The ADSR could mix in the legacy waste with thorium and burn it by using the high range of neutrons produced by the spallation process, reducing its toxicity.

‘Another advantage of the accelerator-driven approach is that you may not need to mix uranium or plutonium with thorium to make it run as a fuel. The neutrons produced by the spallation will do the conversion to U233. At the moment, we are performing calculations on this.’

The only operational reactors currently using thorium are in India, which possesses abundant thorium reserves but little uranium. These are conventional, solid fuel LWRs. Meanwhile, it is expected China will be the first nation to develop a new MSR. In January 2011, the China Academy of Sciences launched a US$350m R&D programme on LFTRs, known locally as the thorium-breeding molten salt reactor (TMSR). The 2MW test unit, developed by the Shanghai Institute of Nuclear Applied Physics, is currently expected to be operational by 2020. However, the programme has been delayed by two years as it is taking longer than expected to train the 700 scientists required.

India is also slowly developing a thorium programme at its Bhabha Atomic Research Centre in Mumbai, although primarily for use in an advanced heavy water reactor rather than MSRs. Meanwhile, Russia is believed to be developing thorium as part of a generic nuclear research programme.

Reborn in the USA?

The USA does not have a concerted thorium programme at present, although there are interesting developments in Cambridge, Massachusetts, home to Transatomic Power. An offshoot of MIT, Transatomic Power is developing a Waste-Annihilating Molten Salt Reactor (WAMSR) designed to run not on thorium, but on the United States’ considerable stockpile of radioactive waste.

The company’s ultimate aim is to produce a 500MW unit. It estimates that it can build such a plant for US$1.7bln, roughly half the cost per megawatt of current LWRs. Transatomic Power, backed by the founder of E Ink, Russ Wilcox, appears to be serious. It has recruited highly experienced nuclear engineers from Westinghouse and MIT scientists, and it recently scooped the top award at the Department of Energy’s 2013 Energy Innovation Summit. So far, it has raised a fraction of the US$200 million needed to build a small prototype. Aside from materials science, the lack of funding – from both the private and public sectors – is likely to continue to thwart ambition for thorium.

Nuclear consultant Steve Kidd, former Deputy Director General of the World Nuclear Association, cannot foresee a bright future for thorium. ‘It's only when the likes of Westinghouse, GE Hitachi and Areva come into the frame that thorium will get going. Most of the current development is by physicists and other scientists at government research centres, and they are not too fussed about commercial development. It may be that the thorium cycle is superior to uranium and using thorium in reactors is perfectly technically feasible, but nobody is going to commercialise it because in the 1950s, we decided to go down the uranium route,’ he says.

The London-based Weinberg Foundation, founded in 2011 to promote thorium energy (and named in honour of Alvin Weinberg), refuses to be downhearted by cynicism in the conventional nuclear industry. Researcher David Martin says, ‘In my opinion, a commercial MSR is quite likely within 20 years. There are many reasons for this, firstly the economics of the current nuclear industry. They don’t have a product they can build on time and to budget. An MSR is basically a chemical set with some surrounding concrete, and it operates at atmospheric pressure so there’s no need for multiple redundant safety systems and high-pressure vessels. In theory, the key components of MSRs could be constructed in a modular fashion, rather than built in situ on a bespoke basis.

‘Furthermore, fuel fabrication is more chemistry than physics. It’s far easier to manufacture the few cubic metres of fluoride or chloride salts and fabricate fuel rods. Given the safety of operation, I just can't see how it could be more expensive than LWRs.’

Martin thinks it’s high time the UK Government funded more research. ‘There should be a vigorous programme to recapture some of the optimism that characterised the first wave of nuclear reactors. In the UK, nuclear R&D is a mess. The fragmentation of the industry was an act of vandalism that has left us without a nuclear research base.’