Resistance is futile

IN 1908 a Dutch physicist, Heike Kamerlingh Onnes, cooled helium gas to below its boiling point of -269°C, or just four degrees above absolute zero (4K). Three years later, exactly a century ago, he observed that when liquid helium was used to chill mercury, the metal's electrical resistance suddenly vanished, allowing current to flow completely unobstructed. He had discovered superconductivity.

The implications seemed nothing short of revolutionary. Perfectly efficient electric cables, more powerful generators and motors, magnetic levitation and a host of other technological wonders beckoned. Since then most of those early hopes have been dashed. A hundred years on, superconductors have found widespread use in just one technology, magnetic resonance imaging (MRI), which lets doctors peer inside patients' bodies. But this may be about to change, as materials which retain their remarkable properties at higher temperatures start to be put to work where Kamerlingh Onnes thought they belonged from the start: in generating and transmitting electricity without resistance.

Electrical resistance arises when the free electrons passing through the rigid ionic lattice of a metal occasionally bump into its constituent ions. The collision transfers energy from the electron to the ion, which starts vibrating more vigorously as a result. In other words, some electrical energy is lost as heat (since the temperature of a substance is a measure of how furiously its atoms are vibrating).

In 1956 Leon Cooper, an American physicist, figured out that electrons in a superconductor avoid this fate by overcoming their mutual repulsion and pairing up. As a negatively charged electron passes through a lattice, the ions along its path feel a slight attractive force and stray as far into the electron's wake as the lattice structure lets them. This distorts the lattice, creating a concentration of positive charge. Other electrons zipping along in the vicinity will then be drawn to this region and, as a result, towards the original electron.

Pair-shaped

Normally, the pull of one passing electron on another is drowned out by the ions' own wriggling. Cool the metal down enough, though, and the wriggling becomes sufficiently weak for one electron's gentle tug to be felt by another and for so-called Cooper pairs to form. Once paired, electrons stop behaving like ordinary particles of matter and, together with other similar pairs, enter a quantum state in which they become oblivious to the ions, and so lose no energy bumping into them. Current can then pass through the lattice without resistance.

The rub is that for Cooper pairs, cool enough means no more than about 30K, or -243°C. The only way to achieve temperatures this low involves the finicky and expensive process of liquefying helium. As a result, low-temperature superconductors are used only in devices where there is no substitute for their remarkable properties. In MRI a more powerful magnet results in a sharper image and a quicker scan. Helium-cooled superconducting coils, typically made of an alloy of niobium and titanium encased in copper cladding, create magnetic fields ten times stronger than similar-size permanent magnets can muster. That is worth paying for, which is why MRI makes up the bulk of the €4.5 billion ($6.1 billion) global market for superconductors, according to the resolutely named Consortium of European Companies Determined to Use Superconductivity, or Conectus (see chart). Siemens, a German engineering giant and a leading maker of such devices, has seen demand for superconducting MRI machines grow at the expense of the non-superconducting sort.

Another niche where powerful superconducting magnets are indispensable is high-energy physics. The Large Hadron Collider (LHC), the world's biggest particle accelerator, uses a staggering 1,200 tonnes of superconducting wire, similar to the sort used in MRI, in order to speed protons up to within a whisker of the speed of light and to collide them inside vast detectors, themselves stuffed with several hundred tonnes of superconducting materials.

Helium-cooled superconductors like those found in MRI machines and the LHC cannot, however, compete with ordinary copper wire in more pedestrian applications, like transmission cables, where the advantages of superconductivity do not merit the enormous costs. But in 1986 Georg Bednorz and Alexander Müller, two researchers at IBM's Zurich laboratory, discovered that an exotic ceramic material behaved like a superconductor at 35K. Because Dr Cooper's pairing theory only works up to about 30K, some other, as yet unexplained mechanism must be at work. Dr Bednorz and Dr Müller's discovery therefore provoked a flurry of research. Soon, physicists were cooking up ceramics that superconducted at around 90K. This may not sound all that balmy, but it is above 77K, the boiling point of nitrogen. Unlike helium, which is extracted from natural gas, nitrogen can be readily harvested from the air, and cooled for a fraction of the price.

Exactly how such high-temperature superconductors (HTSs) work remains a mystery, but that has not stopped engineers from trying to exploit them. On paper, HTSs offer many advantages over conventional copper wires. They can carry five to 20 times more current in the same unit area while reducing the amount of energy lost as heat by 75-97% (depending on whether the current is alternating or direct), even after accounting for all the nitrogen-cooling paraphernalia. Moreover, modern copper-based grid systems tend to be cooled already. This is done either with mineral oils, which present a fire hazard, or with sulphur hexafluoride, the most potent greenhouse gas. If nitrogen leaks out, by contrast, it simply boils off into the air from which it was originally extracted.

The reason HTSs have not taken hold is that brittle ceramics are incredibly tricky to spin into flexible wires, principally because their crystals need to be perfectly aligned in order for resistance to remain low. A decade ago, the only way to do this was to sprinkle the materials into silver tubes, which were then pulled into thin filaments. The resulting wire was up to 80% silver. This was unsustainable even before silver prices went through the roof, says Jack McCall of AMSC (formerly American Superconductor), the world's biggest producer of HTS wires.

In the past few years, however, producers have devised clever manufacturing techniques which use only a tenth as much silver as before. Nowadays AMSC starts by engraving a microscopic pattern onto a sheet of nickel-based metal, to align the crystals. Several buffer layers of non-superconducting material are deposited on top to refine the pattern. Then a coat is applied of yttrium, barium and copper, the metallic elements of YBCO, today's HTS recipe of choice, along with impurities just nanometres (billionths of a metre) across. These impurities help tame the magnetic fields caused by current inside the wire, increasing its capacity. A thin layer of silver comes next. Finally, the whole sandwich is heated in the presence of oxygen, which combines with the precursor metals into fully fledged YBCO.

Techniques like this have helped bring the price down by 90% from 1990s levels, though it remains ten times higher than that of an ordinary copper cable, which sells for $15-25 per kiloamp per metre, the industry's preferred unit. But it is low enough to stoke interest.

New York state has long been at the forefront of HTS pilot projects, with one wrapped up in Albany, another currently running in Long Island and a third being rolled out in New York City. John Love of NYSERDA, an agency charged with revamping the state's power infrastructure, explains that the penchant for superconductivity is born of necessity. New York's dilapidated grid is struggling to keep up with growing demand for electricity from its large, densely populated urban areas. (The state can also tap the local technical nous of SuperPower, a maker of HTS wire, and Brookhaven National Laboratory.)

It is in crowded cities that superconducting cables are likely to take hold first, according to Mark Blamire, who studies superconductor technologies at Cambridge University. In such places, space to lay new cables is scarce and installing new capacity is constrained by regulations or landowners reluctant to see their backyards dug up. Simply rethreading existing infrastructure with superconducting wires could significantly increase the supply of electricity to power-hungry city dwellers.

Catching a second wind

In the longer term many in the industry are looking to the renewable-energy sector as a source of demand for their superconducting wares. Most wind and solar power will be generated in remote places far from where it is consumed. As these sources of power spread, which they are likely to given global commitments to cutting carbon emissions, the electricity they produce will need to be carried over vast distances where power losses due to residual resistance, as much as 7-10% for conventional cables, begin to hurt.

HTSs have a role to play in generating electricity, not just transmitting it. Winding superconducting wires into coils would make it possible to build turbine units that are half the size and weight of conventional ones. According to America's Department of Energy (DoE), a 10% increase in tower height can increase a turbine's energy output by a third. Lighter superconducting turbines could thus be perched atop higher towers while still being capable of generating tens of megawatts of electricity.

“Superconducting cables are likely to take hold first in crowded cities with little space for new cables.”

The principal engineering challenge in all HTS revolving devices, be they wind-turbines, steam generators or motors, is that they must be able to withstand forces of up to 5,000G while maintaining a constant, very low temperature. Moreover, the ceramics must retain their desirable properties even in powerful magnetic fields.

After a successful trial of a 4MW generator, Siemens has teamed up with the Karlsruhe Institute of Technology to design one capable of churning out several hundred megawatts. In February the project won the support of Germany's ministry of economics and technology. Across the Atlantic, the DoE recently awarded a $3.1m grant to the University of Houston and SuperPower and another $1.4m to the Brookhaven lab, working with AMSC, to come up with a cost-effective wire tailored for a wind turbine. The funds came from a $156m kitty for projects to improve America's energy efficiency.

The DoE is also exploring another grid technology: superconducting magnetic energy storage (SMES). Because current flows unobstructed through a superconductor, once it is fed into one, it will continue flowing for a while without the need to expend energy to nudge it along. SMES systems could one day offer an alternative to lead-acid batteries as a way to store electricity and manage loads across smart grids. But existing SMES prototypes can only store energy for a few minutes at a time. ABB, a Swiss-Swedish conglomerate, has received $4.2m from the DoE to lead an effort to extend this to an hour.

“Superconductors have a role to play in generating electricity, not just transmitting it.”

Superconductors might be making their way onto trains, too—though not, as many magnetic-levitation aficionados had hoped, to replace wheels. With money from the European Union's Railenergy project, Siemens is developing an HTS traction transformer that, thanks to its compact size, could fit beneath the train's floor rather than occupy bulky compartments. And, weighing 40% less, it would let trains go faster while using less energy.

But many industry-watchers believe that the HTS technology ripest for commercialisation is fault-current limiters (FCLs). Fault currents are sudden surges of power. They can be caused by a short circuit—when a cable is struck by lightning or hit by a falling tree, say. Power grids are equipped to handle such eventualities but this requires ensuring that all kit linked to them can withstand the spike in the current, often to many times the normal level, before circuit breakers kick in several milliseconds later. The circuitry needed to do this typically increases a grid's total impedance (the alternating-current equivalent of resistance) and reduces its efficiency.

Superconducting FCLs, by contrast, are transparent to electricity until the current surges past a critical level. Then, by dint of another fundamental property of superconductors, they abruptly become resistive again, only to go back to business as usual when the surge dissipates. All this happens in a split second, in effect turning superconductors into reusable fuses which, because they rely on the laws of physics, are fail-safe to boot.

Fault currents are likely to become a growing problem as smart grids grow increasingly complex, with ever more suppliers and complicated load management. Earlier this year Siemens, AMSC and Nexans, another big cable-maker, successfully tested a superconducting FCL at 115 kilovolts (kV), the highest voltage yet, demonstrating that such devices could work on transmission lines. (Distribution lines carry electricity at below 50kV.) A dual-purpose distribution-FCL cable has also been installed as part of a New York HTS pilot called Hydra.

However, these ventures would probably not have got off the ground on strictly commercial grounds, and had government support. Jeff Quiram, the boss of Superconductor Technologies, another cable-maker, pins part of the blame for HTSs' slow adoption on utilities' innate conservatism. But he is not alone in arguing that it is also partly the upshot of deregulation of America's electricity markets, which began in the late 1990s and led to the dismantling of overweening monopolies. Smaller, nimbler firms are a boon to consumers. Often, they are quicker to embrace newfangled technologies. But they may be less willing to stump up the large sums needed for projects with a distant and uncertain break-even point.

Like Mr Quiram, AMSC's Mr McCall is eyeing places like South Korea and China, whose state-backed monopolies have the luxury of a longer time horizon, and appear more willing to take a punt on innovative technologies. In September the Korea Electric Power Corporation, which generates almost 90% of the country's electricity, began sending 50MW of power through superconducting cable at the Icheon substation near Seoul. The cable was produced by LS Cable, a South Korean firm. The two companies have since agreed to deploy 50km of cable in commercial grids by the end of 2015. Some of the first batches will be used in a demonstration smart grid being built in Jeju, an island off the south coast of the Korean peninsula. The Chinese, too, are hopping on the bandwagon. In February Baiyin, a city in Gansu province, got a spanking new substation which sports a full set of superconducting kit: cables, a SMES system and an FCL, as well as a transformer.

Damien Hampshire, editor-in-chief of Superconductor Science and Technology, a specialist peer-reviewed journal, believes that the future of superconductors lies elsewhere still. ITER, a huge experimental fusion reactor being built in France, will use 80,000km of low-temperature, helium-cooled superconducting wire to generate the immense magnetic fields needed to confine the 150 million-degree plasma in which hydrogen atoms will fuse into helium, releasing oodles of energy. When the project is fired up in 2019, its superconducting magnets alone will have accounted for a third of the €15 billion price tag.

Drinking the cool aid?

Cynics quip that fusion is, like superconductors only more so, just a few years away—and always will be, because the energy gleaned from the reactor shows no sign of surpassing the vast amounts needed to keep it going. It is certainly far from assured that such “burn with gain” is feasible. But Dr Hampshire speaks for many physicists when he touts fusion as the ultimate solution to the world's energy conundrum. Whether or not his optimism is justified, superconductivity may yet prove everyone wrong. Its next killer app could well take everyone by surprise, just as MRI did 30 years ago.