¿ìè¶ÌÊÓƵ

One of the most formidable generators of electrical power is gradually taking shape in a basement laboratory in the heart of London’s West End. Drawing less energy from the national grid than a domestic kettle, the generator should soon be able to produce a pulse of power about 30 times larger than all of Britain’s power stations put together.

¿ìè¶ÌÊÓƵs want to use this high-voltage tool to force an enormous current through a fine strand of frozen hydrogen. Their aim is to gain a better understanding of thermonuclear reactions and, ultimately, how controlled nuclear fusion, which has proved remarkably elusive so far, could provide an alternative source of electrical energy. For this, they would replace the hydrogen with its isotopes, deuterium and tritium – fusion’s fuel.

The device, known as a pulsed-power generator, will be the largest of its kind in the civilian world, smaller only than half-a-dozen or so similar devices in military laboratories in the US, Russia and Europe. Known as Magpie (Mega Ampere Generator for Plasma Implosion Experiments) and being built by the plasma physics group at Imperial College, it is a tangible break from the traditional approaches to fusion research.

Magpie will not be ready for use until June, but so keen are its developers to get on with their investigations that they are buying time on Russia’s largest generator, the Angara-5-1, at the Troitsk laboratory of the Kurchatov Institute, once a top-secret military establishment on the outskirts of Moscow. Over the next fortnight, with physicists from France’s National Agency for Scientific Research (CNRS), they will finish a set of experiments started last September in what heralded one of the first collaborations of its kind. The Russian institute is charging $45 000 for its services, a fee that is being met jointly by the British government, CNRS and the University of California at Irvine.

Traditionally, physicists have pursued two routes in trying to fuse deuteron and triton, the nuclei of deuterium and tritium, to produce a helium nucleus and high-energy neutrons. Both methods face the same demands: the reactive particles must be held together at a temperature of 100 million °C long enough for them to fuse. The Lawson Criterion, which defines the minimum requirements for producing fusion energy, says that 1014 particles of deuteron and triton must be held within a 1-metre cube for 106 seconds, or in any combination which achieves the same factor of 10²⁰. (So for a 1-centimetre cube, the figure would be 10¹⁴ particles for 1 second.) Fusion can be achieved by increasing the density of particles and decreasing the time that the particles are kept together; or vice versa. But at the high temperature required, the fuel is an unstable plasma, consisting of positively-charged nuclei in a sea of electrons that have been stripped from them, and controlling the mixture is difficult.

The more popular of the two approaches uses a magnetic field to confine the particles while a large current is applied to raise their temperature. The most promising configuration of magnets is the doughnut-shaped Tokamak design, which is the focus of much fusion research in Britain, Russia, the US and Japan. But success with magnetic confinement fusion, as the approach is known, has been modest, and extremely expensive. Researchers have not been able to contain the comparatively dilute plasma (around 1014 particles, or about 0.001 micrograms, of deuterium per cubic centimetre) in the magnetic field for long enough at the high temperature to produce energy reliably.

The other approach, favoured by military researchers because of its compactness, bypasses the problem of containing an unstable plasma for a long time. Instead, a laser beam or ion beam transfers its energy quickly into a tiny capsule, between 1 and 3 millimetres in diameter, to raise the density of the fuel to around 10²⁵ particles per cubic centimetre. The aim is to heat up the surface of the spherical capsule so quickly that it explodes uniformly, thereby creating a reactive implosion – the same mechanism that drives a jet engine forward – that forces the particles within the capsule into a dense plasma in order to fuse. But this approach, known as inertial confinement fusion, has been even less successful. One of the main constraints is the difficulty of building a high-energy beam powerful enough to ensure fusion. For instance, a laser beam able to provide the necessary 100 kilojoules of energy for 1 nanosecond would be generating 100 terawatts. On the basis of current technology, it would also be the size of two football fields and cost about Pounds sterling 1 billion.

As a result, small groups of plasma physicists around the world have renewed their interest in another approach to fusion research, a technique that was in fact the first method used to observe thermonuclear fusion, during classified experiments in Britain, France, the US and the Soviet Union in the 1950s. The source of particles is a plasma that is denser than a Tokamak’s but less than that used in inertial confinement fusion. Consequently, the power required to raise the temperature of the particles to 100 million °C, and the period of application, lie between the two other methods’ extremes.

The method is known as ‘pinch’ fusion, because the electric current pumped through the plasma generates an encompassing magnetic field that constricts the particles. The first experiments were not self-sustaining (delivering more energy than they absorbed) because the plasma proved too unstable to contain: the hotter it became, the faster it broke up. Subsequent development of the technique as a controlled source of energy was quickly abandoned in favour of the two other methods. But disappointing results from those have regularly persuaded some physicists to reconsider pinch fusion.

The two most significant breakthroughs came over the following two decades. First was the development of pulsed-power generators at the Atomic Weapons Research Establishment at Aldermaston in the mid-1960s; then in the early 1970s, scientists at the Los Alamos National Laboratory in New Mexico fabricated fibres of deuterium cryogenically. The solid fibres yield plasmas with a density of around 10²⁰ particles per cubic centimetre, and can now be made as thin as 10 micrometres across – about one-tenth the thickness of human hair. Because thinner fibres have fewer particles, less power is required to heat them up. What is more, the most advanced generators can pump power into a fibre so quickly that the resulting plasma has no time to become unstable.

‘Well, that’s the theory anyway,’ says Peter Choi, the plasma physicist at Imperial College who helped to raise Pounds sterling 1.1 million for Magpie’s construction. He convinced the Science and Engineering Research Council and the Ministry of Defence that his team could build the device, so saving up to Pounds sterling 5 million.

When researchers use the generator, in Imperial’s basement, a 3-centimetre long frozen fibre will be extruded from a cryogenic mixer directly into a vacuum tube, about 1 metre in diameter and around 30 centimetres high. The tube, which sits with the mixer at the centre of the laboratory, consists of two electrode plates, one at the top and one at the bottom, made of stainless steel and separated by an insulating perspex container where the fibre hangs.

The tube rests on a 2-metre diameter duct that rises 3 metres from the middle of the laboratory floor. The vertical duct is built from two concentric cylinders of stainless steel, separated by about 17 centimetres, to which the vacuum tube’s electrodes are connected. Four similar ducts, about 1 metre in diameter and 3.3 metres long, radiate from the pillar at right angles to each other to large banks of capacitors in the corners of the room.

Each bank contains 24 capacitors, which together store 86 kilojoules of energy drawn from the national grid. They are arranged so that they are connected in parallel for charging and in series for discharging, an arrangement devised in the mid-1920s by the German physicist Erwin Marx. Researchers have nicknamed the Marx modules Chico, Harpo, Groucho and Zeppo.

De-ionised water, which fills the gaps between the concentric cylinders and bathes the perspex of the vacuum tube, is the medium used to transfer the energy from the Marx modules to the fibre. To charge up, Magpie’s capacitors draw just 4 kilowatts of power for 90 seconds. When switches in the modules are flicked, the energy transfers to the horizontal ducts within 1 microsecond, and then on into the vertical duct and vacuum tube where it discharges over 200 nanoseconds. The brief time in which the energy is dissipated – the key to pulsed-power generation – lifts the power of the generator to around 2 terawatts as it pumps up to 2 mega amperes through the fibre.

The greatest challenge is to prevent the fibre from becoming unstable when power is applied and it turns into a plasma. So it is shrouded in a secondary, or shell plasma, which allows the current, and hence the encompassing magnetic field, to build up much more rapidly in the conducting composite of plasma and fibre. Without this plasma girdle sharing the power being directed into the fibre, the fibre’s natural inductance (a property of all conductors) would resist the rapid rise in current.

As the power increases over the 200-nanosecond pulse, the magnetic field forces the shell plasma to ‘pinch’ onto the fibre, which is by then a plasma itself, albeit a dense one. With a deuterium fibre, fusion would begin to occur at this stage. The trick would then be to make the ‘pinch’ continue until the process became one of ‘radiative collapse’, when the collapsing plasma’s density approached that of a dying star (around 10²⁷ particles per cubic centimetre), says Choi. The material would then radiate more energy than it could absorb.

While Magpie is being commissioned, the Angara-5-1 is providing a useful testbed for his team’s ideas. Results from the experiments should give plasma physicists at Imperial a head start when Magpie’s commissioning trials are complete in 1994.