At the going down of the nuclear sun

At the going down of the nuclear sun

INDEX TERMS Nuclear power|fusion reactors, prospects;Scientific research|nuclear fusion, source of power;Energy|nuclear fusion, research and development;Research and development|nuclear fusion, International Thermonuclear Experiment Reactor;

DATE 16-Sep-95

WORDS 3118

Nuclear fusion promises to be an exciting and elegant way to generate power. But it is no more of a panacea than nuclear fission has been
IN LABORATORIES around the world, physicists fuss over dim suns in leaky bottles. Their aim is to make a bottled sun bright enough to serve as an inexhaustible source of energy: a nuclear-fusion reactor. Decades of research have taken them a long way in the direction of this goal, but they are still far from reaching it. Indeed, it sometimes seems to recede as quickly as the scientists progress.
The International Thermonuclear Experiment Reactor (ITER) is meant to end this piecemeal progress with a decisive coup de theatre. A collaboration between all the countries involved in large-scale fusion research, it will have a bottle (actually a complex set of magnetic fields) larger and stronger than any previously built. Within that bottle, ITER's little sun (an electrically charged plasma of rare forms of hydrogen) will be hot enough for fusion reactions between the atoms to sustain themselves for long periods, producing more energy than the reactor consumes.
The reactor now being designed was conceived at a time when enthusiasm for big scientific projects was at a peak; the proposal dates from a Soviet-American summit in Geneva in 1985. Western Europe and Japan quickly signed on, too, and in 1992 the slow business of turning a concept into a practical engineering design got under way. But the final decision about whether actually to build the reactor has not yet been taken, and must now come at a moment when the appetite of governments for big science projects with far-off benefits has waned. As a result, ITER is now in danger.
Can the world afford to turn its back on an exciting new source of power while its energy demands are growing rapidly? In a way, it already has. Fusion's ugly sister, fission, has been producing electricity for decades. But in most industrial countries, fission has fallen on hard times. It is associated with radioactive waste, the proliferation of nuclear weapons, and the looming threat of nuclear cataclysm.
'It's sure death for a politician to say, 'I'd like nuclear energy,' says Cynthia Carter of the advanced-energy division of America's Energy Department. 'We're trying to close these things out - not highlight them.' In the industrial countries little effort is going into the development of new forms of fission, despite the technology's proven capabilities and potential. So why should these same countries want to spend a huge amount of money on a new, unproven form of nuclear power?
Breaking up and getting together

Although both are nuclear reactions, fusion and fission are very different beasts. Small atomic nuclei will give up energy if clumped together. That is fusion. Large nuclei will give up energy if broken apart: that is fission. All elements except iron, which is perfectly happy as it is, can in principle be made to release energy in one or other of these ways. In practice, fusion works best with very light nuclei, fission with very heavy ones.
Some heavy nuclei are unstable. If you hit one with a neutron, it will break up into smaller nuclei and more neutrons. These neutrons can go on to hit other nuclei in turn. Get enough such nuclei close together and the knock-ons will produce a chain reaction. Unfortunately, the heavy elements required - uranium or plutonium, normally - are expensive and rare, and the new nuclei produced are radioactive and dangerous. Worst of all, the chain reaction can spin out of control, leading to a meltdown.
Fission is reasonably simple to start. The first fission reactor, built in a squash court in Chicago under the guidance of Enrico Fermi, worked first time. But - witness Chernobyl - it can be hard to stop. Fusion, on the other hand, is hard to get started. While fissile nuclei are eager to fall apart, those which might fuse are loath to get close enough to do so.
To bring two light nuclei close enough to fuse means overcoming their natural repulsion. They have to be moving fast, which means they have to be hot - hence the term 'thermonuclear' to describe fusion reactors and their untamed relations, hydrogen bombs. Getting things hot enough while keeping them under control is hard. But if you can arrange it, the result will be more attractive than a fission reactor in a number of different ways.
The first has to do with the fuel. First-generation fusion reactors will use a mixture of deuterium and tritium, heavy forms of hydrogen that contain extra neutrons. Deuterium is plentiful in the oceans. Tritium can be manufactured by splitting lithium nuclei in half with neutrons, which is easily done in most sorts of nuclear reactor, and could be done in a fusion reactor. So fuel is not much of a problem. Nor is the exhaust. The end product of a fusion reaction is not a pile of poisonous radioactive muck. It is helium, a harmless (indeed noble) gas.
Fission takes energy out of a lot of fuel at a low rate, a rate which can increase cataclysmically if the chain reaction gets out of control. Fusion takes energy out of a small amount of fuel at a high rate. There is no chain reaction going on and only a little fuel is present in the reaction chamber at any one time. So there is no risk of a runaway disaster. And the only way to get deuterium and tritium to explode in a bomb is by using a fission bomb to trigger the reaction. Whereas a fission reactor can give a would-be proliferator access to all the materials needed for an entry-level fission weapon, a fusion plant is not really any help at all to a weapons manufacturer.
It could help weapon designers, however. Hydrogen bombs use the energy given off by fission weapons to squeeze deuterium and tritium. A scaled-down version of the same sort of effect can be achieved by squeezing the deuterium and tritium with something a little less dramatic, like a set of laser beams. This is inertial-confinement fusion (ICF), and it is a useful tool for researchers who want to know what happens when a bomb goes off. New ICF reactors now under discussion or being designed - notably America's $ 1.1 billion National Ignition Facility and France's Megajoule laser - the megajewel of the FFr10 billion ($ 2 billion) PALEN defence-science project - could get true fusion reactions going, and might provide a technology that could be turned into a power plant. But these are not programmes for international sharing.
Civilian fusion research uses magnetic fields, not barrages of beams, to keep its plasmas in line. Powerful magnetic fields can keep charged nuclei moving in tight spiral orbits; the stronger the field, the more confined this nuclear plasma will be. There are several ways of setting up such fields, but the one fusion workers now prefer is the 'tokamak', a doughnut-shaped arrangement. Inside the chamber, a dollop of hydrogen is heated, giving the nuclei the force to fuse when they slam into each other; the magnets keep the atoms close together, so that such slamming happens frequently.
For a fusion power plant to be worthwhile, the energy of the fusion reactions should keep the plasma hot, and thus keep the reaction going without any external heating. This is called ignition and sustained burn - and it is the main goal of ITER. Once scientists achieve it, all they have to do is keep feeding fuel into the reactor and remove the helium produced. Then they will have a self-sustaining reaction that lasts as long as they can keep the plasma confined.
Mess-up in a bottle

The ITER programme is meant to take up where smaller programmes such as the Tokamak Fusion Test Reactor (TFTR) in America, the Joint European Torus (JET) in Britain and the JT-60 in Japan have left off. These national and regional reactors have taught scientists much about the construction and bottling of suns. But they are simply too small to produce a self-sustaining reaction. Next-generation national and regional programmes would probably still be too small.
To be big enough to be sure of a result, ITER has to be expensive. The idea is to do everything necessary to ensure that the system works, and to provide the information about fusion plasmas needed before smaller, less extravagantly engineered commercial reactors can be designed. ITER would not only be big, but also technologically advanced; to achieve its confinement goals ITER's magnets would be made of new superconductors far more capable than anything made before. Those magnets take up almost 40% of the ITER budget.
A big reactor means a big collaboration, with all the bureaucracy, politics, and confusion that such an undertaking entails. Not least among the problems is the containment and distribution of pork. The result is three design sites: in the United States, Japan, and Germany. At the same time, the pursuit of a single working reactor is bad news for local research programmes. In a world ideally arranged for the benefit of plasma physicists, these would continue in parallel with ITER. This is not that world; international programmes normally mean the sacrifice of national research. That is why the physicist's attitude towards grandiose international approaches in his field often echoes St Augustine's attitude towards virtue: 'Lord, make us truly international, but not yet.'
Scientists are unwilling to shut the old programmes down so long as they produce good results; both JET and TFTR have been extended long beyond their original decommissioning dates. Unfortunately, they are elderly, and there comes a point when they must be replaced; ITER now looks like that replacement. America, for example, will have to forgo the Tokamak Physics Experiment (TPX), a machine designed to exploit recent discoveries such as a magnetic-field configuration pioneered at TFTR called 'reverse shear', which boosts plasma density by as much as a factor of 50. A recent report by the President's Committee of Advisors on Science and Technology (PCAST) called TPX a 'well conceived and innovative advanced tokamak experiment', and said that it would be 'extremely unfortunate' if it were cancelled - a course it then, in effect, recommended.
But fusion's financial problems mean that it faces harder choices than this sort of either/or. There is a real chance of neither. The PCAST committee, headed by John Holdren of the University of California, Berkeley, said that it supported an increase in the fusion-research budget to an average of $ 645m a year between 1995 and 2005. However, it acknowledged that this will not happen. So it based its recommendations on a budget of $ 320m a year.
With that level of funding, the committee saw no likelihood of TPX - at least, not in the near future. And it also saw a need for renegotiating ITER. The new facility would cost about $ 4 billion, roughly a third of the original bill. The smaller ITER would be much less impressive, and would come on-line later; it would be unable to sustain a burn for as long, and it would not be the testbed for new technology originally designed. It would use today's magnets, not tomorrow's.
That would be a big blow for ITER, and the programme's council has rejected the idea. But PCAST's $ 320m fusion research budget is, itself, optimistic. The proposed congressional budget allows a mere $ 229m. The PCAST report envisages a very bleak fusion landscape at this level. There would be ' ..no contribution to an international ignition experiment or materials test facility (ie, no ITER), no TPX, little exploitation of the remaining scientific potential of TFTR, and little sense of progress toward a fusion energy goal.' Ten years after the Geneva summit, those are the realities of fusion research; ITER, as currently envisaged, does not fit. Without America, it may not go ahead at all.
Splitting's heirs

Fusion's future does not, at the moment, look bright. Neither does fission's present. Nuclear fission has not become cheap, let alone 'too cheap to meter'. The costs of building a nuclear power plant are still high. Many countries have given up on them. In Fermi's native Italy there is not a single nuclear power plant, and none is planned. The United States, with 109 nuclear power plants satisfying 20% of its electricity needs, has not built a new one since 1989. The countries where nuclear power has succeeded are technophilic and resource-poor, like France, Japan, and South Korea. It is also to be found throughout the ex-communist world.
Public opinion has soured on nuclear power. A stuck valve at Three Mile Island and a genuine disaster at Chernobyl made it quite clear that a mishandled reactor can become an ecological catastrophe. Furthermore, nuclear waste is a particularly nasty product. Spent control rods, very radioactive and deadly, must be buried or dumped at sea. Though the volume of waste is small, the problem of handling it is not.
Fusion's proponents used to talk about cheap power, too. These days, the emphasis tends to be placed elsewhere. Fusion plants descended from ITER will be technologically intimidating and difficult to engineer; they are not going to compete with gas turbines unless there is no more gas, or limits on carbon-dioxide emission become very strict. It is not clear that they will be appreciably cheaper than fission, either.
It is at this point that fusion's proponents talk of safety and cleanliness. But in neither of these is fusion's advantage as clear-cut as it might seem. The arguments for fusion that seem so obvious in terms of pure physics get cloudier when real engineering is added to the equations.
Fusion produces helium. It also produces neutrons. Magnetic fields have no power over neutrons, so these neutrons, unconfined, will skitter off and strike the wall of the vacuum chamber at a fusion reactor's heart. Over time they will make the wall itself radioactive, and weaken its structure. The wall will therefore have to be replaced regularly. Preliminary plans for a prototype commercial reactor that might follow on from ITER call for its walls to be replaced every two years or so. Each replacement will produce hundreds of tonnes of nuclear waste. Although more exotic materials such as vanadium alloys can alleviate the problem somewhat, they are expensive; ITER's budget may look gold plated, but its walls will be made out of steel.
Fusion's safety is not clear-cut either. Plasmas are tricky, wilful things. Plasma 'disruptions' can vapourise portions of the containment wall - and can do much more serious damage as well. On one occasion, a disruption caused the JET reactor - all 120 tonnes of it - to leap a centimetre into the air. A large disruption could wreck a reactor, though the damage to the environment would be limited, since little fuel is contained at any given time.
However, a bigger environmental risk lurks in the background. The tritium fuel will be created in a blanket of lithium that soaks up some of the reactor's neutrons. Like sodium, lithium is a very reactive metal; it burns or explodes upon contact with moisture. Lithium spiked with high levels of tritium is not to be messed with. A report on fusion by the European Parliament's Scientific and Technological Options Assessment Project describes what might happen if a reactor accident brought the lithium blanket into contact with air or water. 'It will burn with an intense heat, initiating further accident events and itself releasing the tritium contained in the blanket. (A radiation release) on a scale similar to that at Chernobyl can be envisioned.'
Fission reactors, on the other hand, need not be as dangerous in the future as they have been in the past. It is possible to design them in such a way that mechanical failure is far less of an issue. Westinghouse's AP600 is one such reactor. Rather than using mechanical devices, which can fail, it relies on gravity-driven water tanks and air circulation to keep the core cool. It is also possible to design fission reactors where a constant effort is needed to keep the chain reaction going and accidents cause it to shut down, rather than run away with itself.
But it seems that advanced fission research is dying. In the United States, for instance, the Department of Energy's Office of Nuclear Energy expects a mere $ 40m this year to spend on reactor technology. Private industry is interested in incremental improvements of existing designs more than in new technologies. The same is true in many other parts of the world.
Fuse or refuse?

There is no doubt that the world's energy needs are increasing rapidly. Eventually, fossil fuels will become harder to find, and their environmental cost may become unbearable sooner. However, fusion is not the answer to these linked problems - at least not in the short term. Billed as a clean, safe solution to the world's energy problems, fusion is not necessarily much cleaner or safer than fission, and it is a lot less practical.
That does not meant that fusion is not worth pursuing. In the long term, a source of power that uses a ubiquitous fuel - deuterium - will come into its own. There are only a few centuries' worth of coal and uranium around. But given how long a term that is, there is no clear need to pursue fusion at a rush. If the world's main interest is to avoid carbon dioxide, designing better fission plants is a surer short-term bet.
This does not mean that fission is ideal. Nuclear waste is a problem, as is nuclear proliferation. Fission is expensive. But it is available, whereas even its advocates concede that fusion will not be available commercially in much less than 50 years, even if ITER is built.
Aware of this, some proponents argue for fusion in terms of spin-off - the creation of a new superconductor industry is cited as a reason to go ahead with ITER, just as it was invoked to support the now-defunct Superconducting Super Collider. But that argument can be reversed. Relevant technological progress will not stop if ITER stops. Superconducting magnets seem likely to improve because the market wants them. If governments feel like speeding this along, they can do it directly by spending on superconductors. In so doing, they will create a world where, eventually, fusion reactors will be easier to build. The same goes for other relevant areas.
That is why it makes sense to postpone ITER. It would be satisfying to solve the problem once and for all, to stop the incremental process in which Europe builds a better bottle, America a brighter sun, and the final goal stays in the distance. But that satisfaction would be largely intellectual; it would not really solve many problems.
© 1995 The Economist Newspaper Limited. All rights reserved

			At the going down of the nuclear sun

	INDEX TERMS		Nuclear power\|fusion reactors, prospects;Scientific research\|nuclear fusion, source of power;Energy\|nuclear fusion, research and development;Research and development\|nuclear fusion, International Thermonuclear Experiment Reactor;
	DATE		16-Sep-95
	WORDS		3118

			Nuclear fusion promises to be an exciting and elegant way to generate power. But it is no more of a panacea than nuclear fission has been IN LABORATORIES around the world, physicists fuss over dim suns in leaky bottles. Their aim is to make a bottled sun bright enough to serve as an inexhaustible source of energy: a nuclear-fusion reactor. Decades of research have taken them a long way in the direction of this goal, but they are still far from reaching it. Indeed, it sometimes seems to recede as quickly as the scientists progress. The International Thermonuclear Experiment Reactor (ITER) is meant to end this piecemeal progress with a decisive coup de theatre. A collaboration between all the countries involved in large-scale fusion research, it will have a bottle (actually a complex set of magnetic fields) larger and stronger than any previously built. Within that bottle, ITER's little sun (an electrically charged plasma of rare forms of hydrogen) will be hot enough for fusion reactions between the atoms to sustain themselves for long periods, producing more energy than the reactor consumes. The reactor now being designed was conceived at a time when enthusiasm for big scientific projects was at a peak; the proposal dates from a Soviet-American summit in Geneva in 1985. Western Europe and Japan quickly signed on, too, and in 1992 the slow business of turning a concept into a practical engineering design got under way. But the final decision about whether actually to build the reactor has not yet been taken, and must now come at a moment when the appetite of governments for big science projects with far-off benefits has waned. As a result, ITER is now in danger. Can the world afford to turn its back on an exciting new source of power while its energy demands are growing rapidly? In a way, it already has. Fusion's ugly sister, fission, has been producing electricity for decades. But in most industrial countries, fission has fallen on hard times. It is associated with radioactive waste, the proliferation of nuclear weapons, and the looming threat of nuclear cataclysm. 'It's sure death for a politician to say, 'I'd like nuclear energy,' says Cynthia Carter of the advanced-energy division of America's Energy Department. 'We're trying to close these things out - not highlight them.' In the industrial countries little effort is going into the development of new forms of fission, despite the technology's proven capabilities and potential. So why should these same countries want to spend a huge amount of money on a new, unproven form of nuclear power? Breaking up and getting together Although both are nuclear reactions, fusion and fission are very different beasts. Small atomic nuclei will give up energy if clumped together. That is fusion. Large nuclei will give up energy if broken apart: that is fission. All elements except iron, which is perfectly happy as it is, can in principle be made to release energy in one or other of these ways. In practice, fusion works best with very light nuclei, fission with very heavy ones. Some heavy nuclei are unstable. If you hit one with a neutron, it will break up into smaller nuclei and more neutrons. These neutrons can go on to hit other nuclei in turn. Get enough such nuclei close together and the knock-ons will produce a chain reaction. Unfortunately, the heavy elements required - uranium or plutonium, normally - are expensive and rare, and the new nuclei produced are radioactive and dangerous. Worst of all, the chain reaction can spin out of control, leading to a meltdown. Fission is reasonably simple to start. The first fission reactor, built in a squash court in Chicago under the guidance of Enrico Fermi, worked first time. But - witness Chernobyl - it can be hard to stop. Fusion, on the other hand, is hard to get started. While fissile nuclei are eager to fall apart, those which might fuse are loath to get close enough to do so. To bring two light nuclei close enough to fuse means overcoming their natural repulsion. They have to be moving fast, which means they have to be hot - hence the term 'thermonuclear' to describe fusion reactors and their untamed relations, hydrogen bombs. Getting things hot enough while keeping them under control is hard. But if you can arrange it, the result will be more attractive than a fission reactor in a number of different ways. The first has to do with the fuel. First-generation fusion reactors will use a mixture of deuterium and tritium, heavy forms of hydrogen that contain extra neutrons. Deuterium is plentiful in the oceans. Tritium can be manufactured by splitting lithium nuclei in half with neutrons, which is easily done in most sorts of nuclear reactor, and could be done in a fusion reactor. So fuel is not much of a problem. Nor is the exhaust. The end product of a fusion reaction is not a pile of poisonous radioactive muck. It is helium, a harmless (indeed noble) gas. Fission takes energy out of a lot of fuel at a low rate, a rate which can increase cataclysmically if the chain reaction gets out of control. Fusion takes energy out of a small amount of fuel at a high rate. There is no chain reaction going on and only a little fuel is present in the reaction chamber at any one time. So there is no risk of a runaway disaster. And the only way to get deuterium and tritium to explode in a bomb is by using a fission bomb to trigger the reaction. Whereas a fission reactor can give a would-be proliferator access to all the materials needed for an entry-level fission weapon, a fusion plant is not really any help at all to a weapons manufacturer. It could help weapon designers, however. Hydrogen bombs use the energy given off by fission weapons to squeeze deuterium and tritium. A scaled-down version of the same sort of effect can be achieved by squeezing the deuterium and tritium with something a little less dramatic, like a set of laser beams. This is inertial-confinement fusion (ICF), and it is a useful tool for researchers who want to know what happens when a bomb goes off. New ICF reactors now under discussion or being designed - notably America's $ 1.1 billion National Ignition Facility and France's Megajoule laser - the megajewel of the FFr10 billion ($ 2 billion) PALEN defence-science project - could get true fusion reactions going, and might provide a technology that could be turned into a power plant. But these are not programmes for international sharing. Civilian fusion research uses magnetic fields, not barrages of beams, to keep its plasmas in line. Powerful magnetic fields can keep charged nuclei moving in tight spiral orbits; the stronger the field, the more confined this nuclear plasma will be. There are several ways of setting up such fields, but the one fusion workers now prefer is the 'tokamak', a doughnut-shaped arrangement. Inside the chamber, a dollop of hydrogen is heated, giving the nuclei the force to fuse when they slam into each other; the magnets keep the atoms close together, so that such slamming happens frequently. For a fusion power plant to be worthwhile, the energy of the fusion reactions should keep the plasma hot, and thus keep the reaction going without any external heating. This is called ignition and sustained burn - and it is the main goal of ITER. Once scientists achieve it, all they have to do is keep feeding fuel into the reactor and remove the helium produced. Then they will have a self-sustaining reaction that lasts as long as they can keep the plasma confined. Mess-up in a bottle The ITER programme is meant to take up where smaller programmes such as the Tokamak Fusion Test Reactor (TFTR) in America, the Joint European Torus (JET) in Britain and the JT-60 in Japan have left off. These national and regional reactors have taught scientists much about the construction and bottling of suns. But they are simply too small to produce a self-sustaining reaction. Next-generation national and regional programmes would probably still be too small. To be big enough to be sure of a result, ITER has to be expensive. The idea is to do everything necessary to ensure that the system works, and to provide the information about fusion plasmas needed before smaller, less extravagantly engineered commercial reactors can be designed. ITER would not only be big, but also technologically advanced; to achieve its confinement goals ITER's magnets would be made of new superconductors far more capable than anything made before. Those magnets take up almost 40% of the ITER budget. A big reactor means a big collaboration, with all the bureaucracy, politics, and confusion that such an undertaking entails. Not least among the problems is the containment and distribution of pork. The result is three design sites: in the United States, Japan, and Germany. At the same time, the pursuit of a single working reactor is bad news for local research programmes. In a world ideally arranged for the benefit of plasma physicists, these would continue in parallel with ITER. This is not that world; international programmes normally mean the sacrifice of national research. That is why the physicist's attitude towards grandiose international approaches in his field often echoes St Augustine's attitude towards virtue: 'Lord, make us truly international, but not yet.' Scientists are unwilling to shut the old programmes down so long as they produce good results; both JET and TFTR have been extended long beyond their original decommissioning dates. Unfortunately, they are elderly, and there comes a point when they must be replaced; ITER now looks like that replacement. America, for example, will have to forgo the Tokamak Physics Experiment (TPX), a machine designed to exploit recent discoveries such as a magnetic-field configuration pioneered at TFTR called 'reverse shear', which boosts plasma density by as much as a factor of 50. A recent report by the President's Committee of Advisors on Science and Technology (PCAST) called TPX a 'well conceived and innovative advanced tokamak experiment', and said that it would be 'extremely unfortunate' if it were cancelled - a course it then, in effect, recommended. But fusion's financial problems mean that it faces harder choices than this sort of either/or. There is a real chance of neither. The PCAST committee, headed by John Holdren of the University of California, Berkeley, said that it supported an increase in the fusion-research budget to an average of $ 645m a year between 1995 and 2005. However, it acknowledged that this will not happen. So it based its recommendations on a budget of $ 320m a year. With that level of funding, the committee saw no likelihood of TPX - at least, not in the near future. And it also saw a need for renegotiating ITER. The new facility would cost about $ 4 billion, roughly a third of the original bill. The smaller ITER would be much less impressive, and would come on-line later; it would be unable to sustain a burn for as long, and it would not be the testbed for new technology originally designed. It would use today's magnets, not tomorrow's. That would be a big blow for ITER, and the programme's council has rejected the idea. But PCAST's $ 320m fusion research budget is, itself, optimistic. The proposed congressional budget allows a mere $ 229m. The PCAST report envisages a very bleak fusion landscape at this level. There would be ' ..no contribution to an international ignition experiment or materials test facility (ie, no ITER), no TPX, little exploitation of the remaining scientific potential of TFTR, and little sense of progress toward a fusion energy goal.' Ten years after the Geneva summit, those are the realities of fusion research; ITER, as currently envisaged, does not fit. Without America, it may not go ahead at all. Splitting's heirs Fusion's future does not, at the moment, look bright. Neither does fission's present. Nuclear fission has not become cheap, let alone 'too cheap to meter'. The costs of building a nuclear power plant are still high. Many countries have given up on them. In Fermi's native Italy there is not a single nuclear power plant, and none is planned. The United States, with 109 nuclear power plants satisfying 20% of its electricity needs, has not built a new one since 1989. The countries where nuclear power has succeeded are technophilic and resource-poor, like France, Japan, and South Korea. It is also to be found throughout the ex-communist world. Public opinion has soured on nuclear power. A stuck valve at Three Mile Island and a genuine disaster at Chernobyl made it quite clear that a mishandled reactor can become an ecological catastrophe. Furthermore, nuclear waste is a particularly nasty product. Spent control rods, very radioactive and deadly, must be buried or dumped at sea. Though the volume of waste is small, the problem of handling it is not. Fusion's proponents used to talk about cheap power, too. These days, the emphasis tends to be placed elsewhere. Fusion plants descended from ITER will be technologically intimidating and difficult to engineer; they are not going to compete with gas turbines unless there is no more gas, or limits on carbon-dioxide emission become very strict. It is not clear that they will be appreciably cheaper than fission, either. It is at this point that fusion's proponents talk of safety and cleanliness. But in neither of these is fusion's advantage as clear-cut as it might seem. The arguments for fusion that seem so obvious in terms of pure physics get cloudier when real engineering is added to the equations. Fusion produces helium. It also produces neutrons. Magnetic fields have no power over neutrons, so these neutrons, unconfined, will skitter off and strike the wall of the vacuum chamber at a fusion reactor's heart. Over time they will make the wall itself radioactive, and weaken its structure. The wall will therefore have to be replaced regularly. Preliminary plans for a prototype commercial reactor that might follow on from ITER call for its walls to be replaced every two years or so. Each replacement will produce hundreds of tonnes of nuclear waste. Although more exotic materials such as vanadium alloys can alleviate the problem somewhat, they are expensive; ITER's budget may look gold plated, but its walls will be made out of steel. Fusion's safety is not clear-cut either. Plasmas are tricky, wilful things. Plasma 'disruptions' can vapourise portions of the containment wall - and can do much more serious damage as well. On one occasion, a disruption caused the JET reactor - all 120 tonnes of it - to leap a centimetre into the air. A large disruption could wreck a reactor, though the damage to the environment would be limited, since little fuel is contained at any given time. However, a bigger environmental risk lurks in the background. The tritium fuel will be created in a blanket of lithium that soaks up some of the reactor's neutrons. Like sodium, lithium is a very reactive metal; it burns or explodes upon contact with moisture. Lithium spiked with high levels of tritium is not to be messed with. A report on fusion by the European Parliament's Scientific and Technological Options Assessment Project describes what might happen if a reactor accident brought the lithium blanket into contact with air or water. 'It will burn with an intense heat, initiating further accident events and itself releasing the tritium contained in the blanket. (A radiation release) on a scale similar to that at Chernobyl can be envisioned.' Fission reactors, on the other hand, need not be as dangerous in the future as they have been in the past. It is possible to design them in such a way that mechanical failure is far less of an issue. Westinghouse's AP600 is one such reactor. Rather than using mechanical devices, which can fail, it relies on gravity-driven water tanks and air circulation to keep the core cool. It is also possible to design fission reactors where a constant effort is needed to keep the chain reaction going and accidents cause it to shut down, rather than run away with itself. But it seems that advanced fission research is dying. In the United States, for instance, the Department of Energy's Office of Nuclear Energy expects a mere $ 40m this year to spend on reactor technology. Private industry is interested in incremental improvements of existing designs more than in new technologies. The same is true in many other parts of the world. Fuse or refuse? There is no doubt that the world's energy needs are increasing rapidly. Eventually, fossil fuels will become harder to find, and their environmental cost may become unbearable sooner. However, fusion is not the answer to these linked problems - at least not in the short term. Billed as a clean, safe solution to the world's energy problems, fusion is not necessarily much cleaner or safer than fission, and it is a lot less practical. That does not meant that fusion is not worth pursuing. In the long term, a source of power that uses a ubiquitous fuel - deuterium - will come into its own. There are only a few centuries' worth of coal and uranium around. But given how long a term that is, there is no clear need to pursue fusion at a rush. If the world's main interest is to avoid carbon dioxide, designing better fission plants is a surer short-term bet. This does not mean that fission is ideal. Nuclear waste is a problem, as is nuclear proliferation. Fission is expensive. But it is available, whereas even its advocates concede that fusion will not be available commercially in much less than 50 years, even if ITER is built. Aware of this, some proponents argue for fusion in terms of spin-off - the creation of a new superconductor industry is cited as a reason to go ahead with ITER, just as it was invoked to support the now-defunct Superconducting Super Collider. But that argument can be reversed. Relevant technological progress will not stop if ITER stops. Superconducting magnets seem likely to improve because the market wants them. If governments feel like speeding this along, they can do it directly by spending on superconductors. In so doing, they will create a world where, eventually, fusion reactors will be easier to build. The same goes for other relevant areas. That is why it makes sense to postpone ITER. It would be satisfying to solve the problem once and for all, to stop the incremental process in which Europe builds a better bottle, America a brighter sun, and the final goal stays in the distance. But that satisfaction would be largely intellectual; it would not really solve many problems. © 1995 The Economist Newspaper Limited. All rights reserved