快猫短视频

Editorial: 鈥�Competition will fuel the nuclear fusion quest鈥�

Nuclear fusion will cost a fortune 鈥� or will it? A new wave of upstart companies think they鈥檝e found cheaper, quicker ways to build a second sun

A VAST earth platform looms into view above the treetops of Cadarache in France鈥檚 sultry south-east. It measures 1 kilometre long by 400 metres wide, and excavators dotted around it are digging out pits to be filled with massive, earthquake-proof concrete foundations. These foundations need to be strong: 18 giant, supercooled superconducting magnets, each weighing 360 tonnes, will be part of a payload totalling 23,000 tonnes. This is the site of ITER, an international scientific collaboration with funding of 鈧�15 billion.

Meanwhile, in an undistinguished building 9000 kilometres away on an industrial park in Redmond, Washington state, a handful of researchers are gathered around a slender cylindrical apparatus about 16 metres long. There are no massive foundations and no expensive cryogenics. The object of the researchers鈥� interest is smaller than one of ITER鈥檚 magnets.

The disparity in scale is striking, especially when you consider both pieces of kit have the same goal: to harness the awe-inspiring power of nuclear fusion. Which project is more likely to realise fusion鈥檚 promise of clean, nigh-on inexhaustible energy? ITER certainly has the funding and the physics and engineering expertise. It would be most people鈥檚 bet. Yet some diminutive upstarts are now challenging that assumption.

What the newcomers lack in size, they make up in ingenuity and dynamism, their backers say. In Redmond and elsewhere, they have gathered some serious money behind their promise to produce the first commercial fusion reactors within years, not the decades ITER will require. Could an upset be on the cards?

There鈥檚 no secret to our interest in fusion: it is what powers the stars, including our sun. At the hundred-million-degree temperatures that exist in the sun鈥檚 core, the nuclei of light atoms fuse together to form heavier nuclei, liberating colossal amounts of energy 鈥� the energy that illuminates and warms our planet some 150 million kilometres distant. What we wouldn鈥檛 give to tame that power for ourselves.

It鈥檚 not that we haven鈥檛 mastered the basics. Humanity鈥檚 first successful experiment with fusion came on 1 November 1952, with the above the Pacific atoll of Enewetak in the Marshall Islands. That demonstrated two things. First, the energy needed to ignite a fusion reaction is huge: an H-bomb requires a Hiroshima-style atomic bomb to set it off. Second, once the reaction is under way, it is virtually uncontrollable.

The chequered history of fusion since then teaches us to be wary of claims to have tamed it (see 鈥淢ore lows than highs鈥�). The magic 鈥渂reak-even鈥� point, at which a reactor begins to produce more energy than it consumes, remains as elusive as ever. Despite our best efforts, for the past half-century practicable fusion has always been something like a half-century in the future.

Commercial fusion would start with two heavy isotopes of hydrogen 鈥� deuterium and tritium 鈥� compressed and heated to an almost unimaginable degree. Under such conditions, these atoms are stripped of their electrons and exist as a charged gas or 鈥減lasma鈥�. In this state, they can be induced to fuse to make helium nuclei, expelling a neutron in the process. The all-important by-product of this fusion reaction is a truly immense amount of energy: 1 kilogram of fusion fuel can produce as much energy as over 10,000 tonnes of coal. This fuel is relatively easy to come by, too. Unlike the heavy-atom fuels of nuclear fission, which must be expensively mined and refined, deuterium can be extracted from water. Tritium is rare in nature but can be 鈥渂red鈥� by bombarding lithium with neutrons handily supplied by the fusion reaction itself.

Doughnut dreams

That masks a big problem, however: deuterium-tritium fusion only kicks in at temperatures above 45 million degrees. Not only does reaching such temperatures require a lot of energy, but no known material can withstand them once they have been achieved. The ultra-hot, ultra-dense plasma at the heart of a fusion reactor must instead be kept well away from the walls of its container using magnetic fields. Following a trick devised in the Soviet Union in the 1950s, the plasma is generated inside a doughnut or torus-shaped vessel, where encircling magnetic fields keep the plasma spiralling clear of the walls 鈥� a configuration known as a tokamak.

This confinement is not perfect: the plasma has a tendency to expand, cool and leak out, limiting the time during which fusion can occur. The bigger the tokamak, the better the chance of extracting a meaningful amount of energy, since larger magnetic fields hold the plasma at a greater distance, meaning a longer confinement time. The current record holder is the UK-based tokamak, which has been operating at the Culham Science Centre in Oxfordshire since 1983. At 15 metres in diameter and 12 metres high it is no baby, but even JET has never quite reached break-even: its best performance is a 16-megawatt output for a 20-megawatt input. It sustained this for less than 10 seconds.

Break-even is the dream was conceived to realise. With a confinement volume over four times JET鈥檚, ringed by magnetic fields almost three times as strong, it should contain a plasma for several minutes, ultimately producing 10 times as much power as is put in.

This long confinement time brings its own challenges. An elaborate system of gutters is needed to extract from the plasma the helium produced in the reaction, along with other impurities. The neutrons emitted, which are chargeless and so not contained by magnetic fields, bombard the inside wall of the torus, making it radioactive and meaning it must be regularly replaced. These neutrons are also needed to breed the tritium that sustains the reaction, so the walls must be designed in such a way that the neutrons can be captured on lithium to make tritium. The details of how to do this are still being worked out.

And the overall success of the project is by no means guaranteed. 鈥淲e know we can produce plasmas with all the right elements, but when you are operating on this scale there are uncertainties,鈥� says , a senior ITER scientist. Extrapolations from the performance of JET and its predecessors suggest a range of possible outcomes, he says. The most likely is that ITER will work as planned, delivering 10 times break-even energy. Yet there is a chance it might work better 鈥� or produce too little energy to be useful for commercial fusion.

ITER鈥檚 great strength is that it builds on the well-established, well-tested physics of tokamaks. But it comes at a huge financial cost and is developing at a snail鈥檚 pace. Even if all goes according to plan, the reactor will not produce its first plasma until 2019, and break-even is not expected until 2026 at the earliest.

鈥淚TER builds on well-established, well-tested physics 鈥� but it comes at huge cost and a snail鈥檚 pace of development鈥�

The goal of ITER鈥檚 new rivals is to reach that point more quickly 鈥� and far more cheaply. The Redmond device, dubbed the , is the brainchild of a company called , and relies on a very different method of establishing and confining plasmas known as a field-reversed configuration. Discovered at the US Naval Research Laboratory in Washington DC in 1960, this process involves accelerating two small, compact balls of plasma into one another at a speed of hundreds of kilometres a second. The conditions created by the collision should, in theory, be sufficient to force the nuclei together, heat them and ignite fusion.

This method has some notable advantages. Although magnetic fields are still used to confine the plasmas, the arrangement is far less elaborate than a tokamak, so the device can be a lot smaller. The reaction is intense and is over in a fraction of a second, and neutrons are only produced at the point where the plasmas collide, making it easy to collect them to breed tritium.

In a peer-reviewed paper published in April this year, Helion researchers show how they used the technique to smash two plasmas together and achieve a temperature of 25 million degrees. That鈥檚 still well below what is needed to ignite fusion, but the team also published calculations showing that ignition 鈥� and even break-even 鈥� should be possible in a device just three times the size of their prototype ().

鈥淲e know it works with small high-density plasmas, and big low-density plasmas,鈥� says , Helion鈥檚 president. 鈥淚t should work with big high-density plasma.鈥� The company, which has already received something like $5 million in funding from NASA and the US Department of Defense among others, is now looking for $20 million from private investors to build what it says could be a commercially viable reactor.

Also pursuing the dream is the Canadian firm based in Burnaby, British Columbia, using a method called magnetised plasma fusion. This set-up also emerged from the US Naval Research Laboratory, this time in the late 1970s. It involves igniting fusion in a plasma violently compressed created in a spinning sphere of liquid metal.

According to company spokesman , the first laboratory tests of the design have gone well, achieving a temperature of 5 million degrees for 1 microsecond. It remains to be seen whether this approach can be scaled up all the way to fusion 鈥� and beyond that to break-even. 鈥淭here are no magnetised plasma experiments that we are aware of at the plasma temperatures and densities necessary for net-gain fusion,鈥� Delage says. 鈥淭he only way to verify this is by experiment.鈥� The firm has raised the $30 million it says it needs, some of it from Amazon founder Jeff Bezos.

Utopian insanity

Tri Alpha Energy, a secretive California-based company, is believed to have raised $90 million for its variant of the field-reversed technique; among its investors is Microsoft co-founder Paul Allen. In a rare public communication a year ago, Tri Alpha researchers showed how they had collided two plasma balls at a temperature over 5 million degrees and held them together for up to 2 milliseconds (). Tri Alpha says it will produce a working commercial reactor some time between 2015 and 2020 鈥� possibly before ITER fires up for the first time.

Those are big claims, and a degree of scepticism would seem to be in order. And indeed you don鈥檛 need to look far to find detractors. Alarm:clock, a website that tracks tech start-ups, has the investment the fusion companies have attracted as 鈥渢estament to the easy-money utopian green insanity which has gripped the imaginations of some of our best venture capitalists鈥�.

But 鈥� and here鈥檚 the surprise 鈥� it鈥檚 hard to find anyone in the know with anything bad to say about the physics behind the new reactors. Even ITER scientists admit that the technology is credible and superficially attractive, if still immature. 鈥淭he tokamak is fairly complicated; some other approaches appear simpler and that appeals,鈥� says Campbell. 鈥淭hey look like a more direct route to fusion.鈥�

He鈥檚 still not about to down tools and switch sides. The body of knowledge built up over the years means a tokamak is still the safest bet to reach break-even first, he says. 鈥淚t may look like there鈥檚 a neat idea waiting to come out, but experience shows there鈥檚 always a catch.鈥�

Unsurprisingly, Delage has a different view. 鈥淚t is worthwhile comparing not just the amount of past experimental work on each technology, but also the work remaining to commercialise it,鈥� he says. Even if ITER does reach break-even in 2026, it will have produced just heat, not the ultimate aim, electricity. More work will be needed to hook it up to a generator. 鈥淔or ITER and tokamaks in general, commercialisation remains several decades away,鈥� says Delage.

The simplicity and smaller size of fusion reactors based on the new technologies 鈥� the companies are aiming for something on the 100-megawatt scale, rather than the gigawatts that are ITER鈥檚 ultimate goal 鈥� could be their great advantage. 鈥淚t鈥檚 a size that allows for factory construction of systems rather than site-specific designs,鈥� says Delage. Wallace agrees. 鈥淚TER is not the sort of thing you could easily roll out in, say, Nigeria 鈥� but we can go anywhere,鈥� he says.

That is for the future. Wallace thinks the new machines might take off first not for power generation, but as neutron sources that could be used to 鈥�transmute鈥� the highly radioactive waste from today鈥檚 fission reactors into low-level isotopes and nuclear fuel. He estimates that 50 Fusion Engines of the size Helion is planning to build could within 20 years eliminate all the waste the US now has stockpiled. Once they are established as neutron sources that just happen to produce power, the small reactors could evolve into commercial power plants, he says.

Such claims have yet to be tested. Some see in the story of ITER and its smaller rivals a potential parallel with the , a billion-dollar investment by the British and French governments. Two decades in the making, Concorde was a technological triumph. Yet it was its less ambitious but more nimble and economical contemporary, the Boeing 737, that became the backbone of air travel. Whether history repeats itself or not, the race to exploit fusion鈥檚 amazing potential is hotting up once more.