Nuclear energy

Nuclear energy will not run cars and trucks on motorways until the much-hyped Hydrogen Economy is a reality. Despite this, urban electric vehicles are just about practical at the time of writing. There is no reason why their batteries cannot be charged from nuclear generated power.

It takes between 10 and 25 years to build a new nuclear plant from the start of the planning process. There's another issue too: the number of science and engineering students with the skills needed to replace retiring reactor workers is below the replacement level in the US. If this continues, staff shortages will restrict new construction or even cause existing plants to close. Energy consultant Ian Fells made the following points at an energy conference in Rimini during October 2005:

Taking a global view, he indicated that even the limited amount of current construction is stretching available capacity to its limits.
- New Scientist, 22 April 2006

Taken together, this means that new plant is likely to take even longer to build than currently estimated and the possibility of finding qualified people to build and run it looks increasingly problematic. In consequence I cannot imagine how nuclear energy can stave off the looming oil crisis, though its now being claimed that nuclear energy was never considered as a short-term fossil fuel replacement. Hmm, that's not what I remember Tony Blair saying in the autumn of 2005. I also see this as another nail in the coffin of the Hydrogen Economy; I cannot see any other non-fossil energy source that, in the short term, could produce enough hydrogen or synthetic fuel to replace the oil used for transport.

Fission power

No new nuclear generators are being planned in the United Kingdom or the USA. However, in June, 2005 both Blair and Bush spoke in support of major new nuclear generation plant programs. This seems totally misguided in the light of information I found in a letter to New Scientist by Peter Shepherd in the 14 May 2005 issue, 2499.

He quotes Storm van Leeuwen & Smith who point out that fission plants are not a good way to go for a no-oil, low carbon future. Provided that high-grade uranium ores are available, nuclear generation produces one third as much carbon dioxide as gas-fired thermal generation, but these ores are relatively scarce. They estimate that the high-grade ores would be exhausted within four years if they were used to produce the entire electrical energy demands of the world. They go on to say that using the remaining poorer ores in fission reactors would produce more carbon dioxide than burning the fossil fuels directly. In other words, mining uranium ore and extracting uranium oxide, "yellow cake", from it uses a lot of oil which we will not have in a decade or two. This assumes that the current uranium cycle (mine, refine, burn, discard) continues until all the economically extractable uranium is used up, something that seems quite likely over a similar timescale to the approaching oil crisis.

The current consensus is that uranium reserves will be exhausted by 2050-2060 at current usage rates. Further, only about 50% of current consumption is newly mined uranium: the rest is the result of recycling nuclear weapons and reprocessing reactor waste.

Fast-breeder reactors have been quoted as a way round this limitation. However, there are a few problems with them: the technology is not mature and so not deployable for some time and there are problems in adequately safeguarding the plutonium fuel from misuse. Fast-breeders burn plutonium to generate electricity while transmuting a U238 blanket into more plutonium which is recycled. U238 is 99.31% of refined uranium metal and cannot be burnt in a uranium-fuelled reactor; only the remaining 0.29% is U235 and this is the only natural isotope that can run a uranium fission reactor. In addition there seems to be some doubt that a fast-breeder could produce enough plutonium to refuel itself, let alone produce the excess needed to cover increasing demand. America, France, Germany and the UK have all abandoned work on these reactors: all considered they look like a non-starter.

Peter Shepherd also excerpted a report from the American Institute for Energy and Environmental Research:

In order to fuel one thousand 1000 Megawatt nuclear plants (a common reference case in many nuclear growth scenarios), a global uranium enrichment capacity roughly 9-10 times greater than that currently operating in the USA would be required. If just one percent of that capacity was instead used to manufacture highly enriched uranium (HEU), then enough HEU could be produced every year to make between 175 and 310 nuclear weapons. With an expanded trade in the specialised materials required to build and operate gas centrifuge and other enrichment plants...diversion of supposedly 'peaceful' technologies will become harder to identify.
- source:

The quoted 1% diversion rate is typical of the inventory discrepancies discovered by International Atomic Energy Authority (IAEA) audits of existing uranium enrichment and reprocessing plants: enough HEU, a.k.a weapons-grade uranium, for just one or two bombs a year could diverted by some clandestine organisation without anybody ever knowing until the bombs were assembled and used. The implication of the letter is clear: not only are fission plants not the way to a low-carbon future but in the current international climate a move to a nuclear future could be rather dangerous.

All that said, if we do have to have a nuclear future, we should take a good, hard look at the Canadian CANDU reactor. This is a natural uranium, heavy water moderated design that allows considerable refurbishment and refuelling during full power operation. It can also handle a wide variety of fuel, ranging from natural uranium through varying degrees of enrichment including spent Pressurised Water Reactor fuel, MOX and degraded bomb material. It is claimed that it can run in slow-neutron breeder mode using the Thorium/U233 fuel cycle, though this doesn't seem to have been demonstrated yet and is hard to imagine, since all experimental Th/U233 reactors so far have been Molten Salt Reactors operating at around 1500 C.

Meanwhile there is still the problem of radioactive waste disposal. The nuclear industry claims that it is just "an engineering problem". Well, maybe so, but if it is just engineering then why, after half a century of nuclear power, has the industry not solved it yet?

Finally, here is something worth thinking about and then asking why we cannot do the same: Germany plans to phase out nuclear generation entirely by 2025 and is on target to produce half of its energy needs from renewables by 2050 in spite of dropping the nuclear option.
- source: New Scientist editorial, 14 May 2005.

Fusion power

This section starts by taking out sun as a benchmark and then looks at the types of fusion generators that have some demonstrable output.

What our sun can do

The only natural fusion power sources we know are stars, so lets look at the sun as a benchmark. Its easy to calculate the Sun's output energy density. Here are its vital statistics:

Diameter 1.392x109 m
Surface area 6.088x1018 m2
Volume 1.412x1027m3
Intensity 2.009x107 W m-2 sr-1
Total output 1.223x1026 W
Energy density 2.721x10-1 W m-3

The Intensity is the energy radiated by the sun perpendicular to its surface. All values shown above except the energy density came from Nasa's "Sun:facts and figures" publication. The energy density was calculated as (surface area * intensity) / volume.

The sun's energy density turns out to be surprisingly low, only about 0.272 watts/cubic metre.

A 1000 MW fusion reactor operating at the sun's energy density would require a working volume of 3,675 million cubic metres, which is a sphere 1.9 km in diameter. This indicates that a fusion generator producing usable amounts of power must either be enormous or it must operate stably at an energy density of at least 100,000 times that of the sun. Even at this energy density the plant would be big. The working volume would be around 37,000 cubic metres, which is a 41 m diameter sphere.

Tokamak - type machines

Currently almost all fusion research is directed at Tokamak machines, which are are large and very expensive devices that seek to generate power using a nuclear reaction that fuses deuterium and tritium to form helium. They achieve this by confining a deuterium/tritium plasma within a toroidal magnetic field where it is heated by injecting energy in the form of electro-magnetic energy (RF induction), inducing a current in the conductive plasma, injecting a high energy beam of ions, or some combination of all three until it reaches a sufficiently high temperature and density for the reaction to occur. The challenge is to keep the plasma hot, stable and reacting for long enough for the fusion reaction to output significantly more energy than was input to establish the reaction. Additional problems are that deuterium is relatively rare and hard to separate from hydrogen, tritium is a radioactive that can only be made in a nuclear reactor and the reaction emits neutrons which make the Tokamak structure radioactive and embrittle steel components of its structure.

No fusion plants are even in the preliminary planning stage. Despite promises over the last 60 years that Tokamaks will deliver plentiful, clean, cheap fusion power once the next machine is complete, it has remained the same distance in the future over all this time. Its rather like a mirage in the desert: the faster you run toward it, the faster it retreats. In reality, even if a clean, safe fusion plant could be built, and its not even on the horizon right now, do you really think the average NIMBY[2] could tell it from a fission plant and would let one be built next door?

Here's a timescale for Tokamak-based development:

Date Project Name Status
1957-61 ZETA Zero Energy Thermonuclear Assembly This was the final machine in the British Z-pinch series, which failed to achieve fusion.
1973-2019 JET Joint European Torus In 2016 this was the largest operational Tokamak magnetic confinement plasma physics experiment. It achived fusion in 1991 and a 16MW sub-second thermal energy output pulse in 1997, but the power input was 24MW, a 65% ratio of input to output (Q=0.65). It can currently maintain fusion for a few seconds at a time but cannot achieve break-even (Q=1). JET's output is entirely thermal energy and has no ability to generate electricity. From 2011 to 2017 it is being used to develop technology for use in ITER. In 2018 JET is expected to run fusion tests using Deuterium-Tritium plasmas in an effort to get closer to break-even (Q=1).
201O-2035? ITER Latin - "The Way" When finished, it will be the world's largest magnetic confinement plasma physics experiment implemented with the world's largest experimental tokamak nuclear fusion reactor. Its aim is to maintain fusion for minutes at a time and to output ten times as much energy as is put in (Q=10). The original plan was to achieve fusion in 2017 but this is now unlikely before 2025 with break-even (Q=1) a few years after that. It is also intended that it will produce all its own Tritium by irradiating a blanket of Lithium that surrounds the Tokamak torus. Like JET, ITER's output is purely thermal energy with no attempt being made to generate electricity. As of 2016 it has a projected 425% cost overrun.
2024-2048 DEMO DEMOnstration Power Plant This is a proposed experimental nuclear fusion power plant, with objectives falling somewhere between those of ITER and a "first of a kind" commercial station. The aim is to produce 2-4GW of thermal power on a continuous basis with a Q value of 25 and to be self-sufficient for Tritium. To achieve this it needs to be about 15% bigger than ITER (linear dimensions, 50% bigger volume) and achieve a plasma density 30% greater than ITER. It is not intended to demonstrate electricity generation before 2048.
2050-? PROTO Prototype Power Plant This is the sucessor to the ITER and DEMO projects. It would be a prototype power station.

Bottom line: even it all goes to plan, don't expect to see any Tokamak-generated power on the grid before 2065-2070 unless it gets a Manhattan Project-like priority.


These are superficially similar to Tokamacs and would generate power using a similar principle: confine a sufficiently hot, dense Helium plasma while the nuclei fuse and extract the energy released as heat, which is used to drive steam turbines. The difference is in the way the plasma is confined. Tokamak's confine the plasma in a torus, which is inherently unstable because of the different path lengths followed by nucleii at different radii within the torus. The stellarator equalises the path lengths by flattening and twisting the confined plasma, which results in a more stable plasma, as predicted by computer simulators. and confirmed on the 10th of December 2015 by the Wendelstein_7-X device and measurements made on the earlier HSX.

Also, unlike Tokamaks and reversed-field pinches, they don't require a toroidal current to confine and heat the plasma and so are more inherently simpler as well as offering longer confinement times. As a result, Stellarators promise to be smaller and cheaper to build than Tokamacs.


A fusor is, in principle, a smaller, simpler device than a Tokamak. It is designed to fuse ions that are both accelerated and confined using high voltage electric fields, an approach known as inertial electrostatic confinement. A detailed description of the origins of this approach, the people involved and of progress so far can be found at

Deuterium fusors are already in use as portable, controllable neutron sources, thus proving that they are practical devices that are capable of steady-state operation. As power sources they can also be fueled with a deuterium-tritium mix, though both fuels have the disadvantages that deuterium is relatively rare and hard to separate from hydrogen, tritium is a radioactive that can only be made in a nuclear reactor or by bombarding lithium with neutrons. The neutrons emitted by deuterium/tritium have the side effect of making the fusor structure radioactive and would make steel components brittle, however calculations show that, if the reaction area is surrounded by a lithium blanket, the fusor should be able to provide its own tritium requirements.

A more attractive approach may be to use the p-B (proton-Boron) reaction. Both protons (ionised hydrogen) and Boron11 are common elements: a fusor using this reaction produces no radiation: its only 'exhaust' is helium, which is a useful product in its own right.

Early fusors, and those used as small neutron sources, are of the Farnsworth-Hirsch type. The Polywell type, originally developed by Dr. Robert W. Bussard, seems to be a better design for electricity generating fusors.

Field-reversed Configuration

Machines based on the field-reversed configuration (FRC) are an offshoot from fusors.

Magnetised Target Fusion

This method, currently under development by General Fusion in Canada, involves creating a deuterium/tritium plasma torus in a vortex cavity at the centre of a spinning sphere of liquid lead. This is then rapidly raised to fusion temperature and density by acoustic compression: the technique is similar to the implosion ignition of a plutonium bomb and has similar timing and symmetry requirements. The fusion energy is absorbed by the lead, which is circulated through a steam generator, which drives a conventional turbo-generator. The lead contains dissolved lithium: some of this is transmuted to tritium, which is separated and used as a feedstock for plasma production. A typical plant would have a 100 MW electrical output if pulsed once a second and the ability to modulate the pulse rate to adjust its output. Current plans are to build a full scale plant and demonstrate a net gain of energy before 2014, followed by commercialisation before 2020. The website gives a clear explanation of the plant's operating principles.