A Critical Look at Wireless Power

The oystermen must have been puzzled by what they saw going up near the north shore of Long Island in New York state at the turn of the 20th century. Nikola Tesla, a Serbian immigrant from what is now Croatia, had set up a laboratory in the midst of potato fields and erected behind it a tower almost 60 meters (200 feet) tall. Tesla said he intended to use it to communicate wirelessly around the world.

The idea was ambitious but not unreasonable, Guglielmo Marconi having just sent the first tentative signals across the Atlantic in 1901. But the Long Islanders living next to the huge tower would have been more shocked—perhaps literally—if Tesla had carried out his second plan, which was so audacious he hid it initially even from J.P. Morgan, the financier who was bankrolling the operation. Tesla wanted to use the tower for wirelessly transmitting not just signals but also useful amounts of electrical power. His strategy for accomplishing that feat was vague, but it seems he had notions of sending power wirelessly to such things as airships in flight and automobiles on the move.

Tesla never did quite finish the enormous tower—Morgan got fed up and cut off funds. Tesla abandoned the lab, which fell into disrepair, and in 1917 the tower was unceremoniously demolished.

Decades later, Tesla’s laboratory was turned into a factory for photographic paper, an operation that left enough toxic waste on the grounds to have the property qualify as a Superfund site. Today the elegant brick lab building is abandoned once again. Plywood covers its windows. All that remains of the tower is its huge octagonal footing, now overgrown with trees.

Scientists and engineers these days, of course, appreciate Tesla’s enormous if rather quirky brilliance—he invented the induction motor, for one thing, and he championed alternating current when Thomas Edison would have none of it, for another. So it’s no surprise that some of Tesla’s admirers are seeking to preserve his old laboratory.

“To think that Tesla walked here, walked on this ground, and had his vision and his dream,” says Jane Alcorn as she surveys the laboratory. Alcorn heads a not-for-profit group dedicated to turning the now-derelict site into a science museum. She knows full well what great things Tesla imagined for the place. “The transmission of power without wires will very soon create an industrial revolution and such as the world has never seen before,” Tesla wrote in a 1906 letter to George Westinghouse.

While Tesla’s vision may appear ludicrous today, things were different before power poles and high-tension lines became a part of the industrialized landscape. Tesla had demonstrated his ability to send power wirelessly over modest distances at a laboratory he had set up earlier in Colorado Springs. But he failed in his grand attempt to scale up the effort. And by the time his power tower came crashing down, people were too busy stringing electrical cables to worry very much about how to do away with them. Recently, though, some hardheaded scientists and engineers have been thinking very carefully about how to do just that.

In 2006, exactly 100 years after Tesla laid off his employees on Long Island, another immigrant from Croatia surprised America with a proposal for sending power through the air. Physicist Marin Soljacic, along with several of his colleagues at MIT, performed a theoretical analysis of a system for projecting useful amounts of power wirelessly using electromagnetic induction, a phenomenon that’s been well known since Michael Faraday first described it in the early 19th century.

In 2007, Soljacic’s team went further and published an article in the prestigious journal Science describing hardware that could light up a 60-watt incandescent lamp using power transferred between two coils separated by a little more than 2 meters. Images of that bulb spookily lit up from afar sparked considerable buzz in the press. But the physics at work was really not so very different from what goes on in any electrical transformer. There, AC current flowing in one coil of wire, the primary, creates an oscillating magnetic field, which in turn induces an AC voltage in another coil, the secondary. In a typical transformer, the magnetic lines of force that link the primary and secondary are channeled through iron, maintaining a tight coupling that keeps power losses to a minimum. If you separate the primary and secondary coils by a distance that’s filled with nothing but air, those losses mount and the transfer becomes inefficient.

“Resonance enables efficient energy transfer,” says Soljacic, describing the basic strategy his team used to get significant amounts of power to flow. It’s not a new idea: Tesla’s eponymous coils use that very same principle.

A good way to understand why resonance helps is to imagine the mechanical analogue. Suppose you wanted to transfer mechanical energy across a room, but all you had coupling the power source with the load was a long and very weak spring. You’d have to pump the end of the spring you were holding vigorously, moving it back and forth as fast and as far as you could until sweat poured down your brow. It wouldn’t be very efficient, but only with such effort would the far end of the spring wiggle a bit.

To make life easier, you could attach your end of the spring to a pendulum swinging in a wide arc, for instance. Now your arm wouldn’t hurt so much, and the far end of the spring would still wiggle. But another difficulty would appear when you tried to attach that far end of the spring to a mechanical load. If you weren’t careful, you’d find the waves of energy being sent down the spring weren’t being absorbed—most of what little energy that got to the far end would just bounce back. To solve this new problem, you could attach the far end of the spring to a second pendulum, one that was built exactly like the first. Now, all you would need to do is give the first pendulum some gentle rhythmic shoves until the amplitude of its swinging became large enough to get the far end of the spring wiggling in time with it. And those little wiggles would in turn have the right timing to get the second pendulum swinging. Despite having only a weak spring as the conduit, you would have transferred power across the room. Then you could do something useful with it—maybe smash a window.

This mechanical analogy may seem a bit loopy, but in fact it provides a very good parallel for what goes on between the coupled resonant electrical oscillators used to transfer power inductively. The mechanical version even shows some of the subtleties of wireless-power systems—for example, that the coupling between the primary and secondary oscillators gives rise to a second, higher frequency of resonance. More important, this thought experiment helps to illustrate a fundamental challenge: As the frequency and amplitude of the oscillations increase, the primary starts to experience significant losses of power. Air resistance would sap the energy of a swinging pendulum, to take one example. For electrical oscillators, most of the losses arise just from the resistance of the wires.

So when Soljacic calls his system “efficient,” he’s speaking in relative terms. The actual plug-to-bulb efficiency in his demonstration would make an environmentalist cringe—it was only 15 percent. Nevertheless, Soljacic and his colleagues were so enthusiastic about the prospects of using such inductive systems to charge cellphones and laptops at a distance that they founded a start-up to commercialize the technology. Dubbed WiTricity Corp., the company, which is located in Watertown, Mass., now has about 20 employees.

Strangely enough, even before Soljacic’s work appeared in print, others at MIT had been looking into the problem of how to send power wirelessly over short distances. Jeff Lieberman, then a graduate student in MIT’s Media Lab (and now the host of the Discovery Channel show “Time Warp”), wasn’t aiming for anything practical; he merely wanted to create an intriguing piece of art—a levitating lightbulb that lit up. So he confronted the same challenge that Soljacic and his colleagues faced: getting power to something without attaching wires to it. Lieberman’s solution differed from Soljacic’s in detail, but the fundamental approach he took was the same.

“I knew the principles of inductively coupled resonant systems had been around,” says Lieberman, who went as far as consulting notes that Tesla had recorded at his Colorado Springs laboratory. After much experimentation, and with the help of Joseph Stark III, another MIT graduate student who wrote his master’s thesis on wireless power transmission, Lieberman succeeded in building a floating, glowing bulb in 2005. But he remained unaware of Soljacic’s work in the physics department until the professor’s scholarly articles on the subject came out.

Those articles brought Soljacic’s group, and later their spin-off company, much attention outside of MIT. And why not? The name WiTricity is, of course, a play on Wi-Fi. It brings to mind images of transmitting power wirelessly throughout homes and offices in the same way that Wi-Fi radios spread Internet access around. A page on WiTricity’s Web site certainly plays on that image, showing an artist’s rendering of a compact transmitting coil positioned in a living-room ceiling, sending power to a wall-mounted television, various lamps, and a laptop poised on a coffee table.

Such thoughts also got Intel to explore the prospects for wireless power at its Seattle labs, where researchers developed something they call the Wireless Resonant Energy Link. “We were inspired by the MIT paper,” says Joshua R. Smith, the Intel engineer who heads the effort. But the system Intel first demonstrated in 2008 uses flat coils rather than the corkscrew coils Soljacic and his coworkers had described the year before. “We didn’t want to use a helix, which is hard to fit into a laptop,” says Smith.

It turns out they didn’t have to. The underlying reason for Soljacic’s helical coils is that they provide not only inductance but also capacitance by virtue of the separation between their adjacent loops. So at radio frequencies these coils resonated without separate capacitors. The advantage, Soljacic figured when he designed them, was that there would be no losses from capacitors, a worry that may have been exaggerated. WiTricity, like Intel, has moved away from Soljacic’s original helical coil design, because the bulky corkscrew form would be awkwardly constraining. “If you’re going to power a device, you want the capture coil to be in the device,” says WiTricity’s chief technical officer, Katie Hall.

Another constraint appears to be less well known, however—or at least it’s less discussed. It comes from the strength of the electromagnetic fields being generated and how they compare with the levels that people would willingly expose themselves to.

How electromagnetic fields affect health is a rich subject, both for what is known about it and what isn’t. The latter category includes many possibilities that may seem more or less reasonable, depending on your perspective. But at radio frequencies, some of the effects are indisputable.

“Exposure to very high RF energy will heat you—there’s no question about that,” says Richard Strickland, who runs the consultancy RF Safety Solutions. He points out that RF exposure guidelines differ from place to place. In the United States, for example, many follow IEEE’s C95.1 standard, whereas Europeans generally adhere to the somewhat stricter guidelines of the International Commission on NonIonizing Radiation Protection (ICNIRP).

Consideration of those limits should be sobering to anyone hoping to send significant amounts of power using electromagnetic fields. Take the ICNIRP guidelines for RF fields at 10 megahertz, the frequency of the system Soljacic and his MIT colleagues built. For this frequency, those guidelines indicate that the general public should not be exposed to magnetic fields in excess of 0.073 ampere per meter, or to electric fields greater than 28 volts per meter. Were this RF energy radiated from a distant antenna, you could apply either the magnetic or the electric-field limit alone, because the ratio of the fields would be a fixed quantity. But inductive power transfer of this kind takes place in what is known as the near field of the antenna (the coil), so the relationship between the electric and magnetic fields is not so simple.

According to their 2007 Science paper, Soljacic and his colleagues measured a magnetic field of 1 A/m at the halfway point between transmitting and receiving coils—almost 14 times the ICNIRP limit. The electric field was 210 V/m, which tops the ICNIRP limit by a factor of 7.5. Things get even worse if you consider the fields closer to the coils. Twenty centimeters away, the magnetic field was more than 100 times—and the electric field 50 times—the ICNIRP limit.

“They are not going to be able to fill a room with fields and not come up with issues,” says Grant Covic, a professor in the department of electrical and computer engineering at the University of Auckland, in New Zealand. He should know: Covic and his Auckland faculty colleague John Boys have been working for two decades on the engineering of such systems, which, despite having garnered little publicity, are in fact used widely for a variety of applications where power cords would be problematic.

One such application is materials handling. Daifuku Co., based in Osaka, Japan, for example, has licensed patents from the University of Auckland to build tracked conveyor systems with moving platforms that are powered wirelessly. These systems, which make up a significant fraction of Daifuku’s US $2.5 billion in yearly sales, don’t generate the fine particles that would contaminate sensitive processes like chipmaking, as brushed electrical contacts tend to do. Such conveyances are useful in other industrial settings, too. Audi and BMW, for example, both use inductively powered carts on their assembly lines, these systems proving more robust than ones that rely on brushed contacts.

Another well-established application is the charging of electric vehicles. More than a decade ago, GM’s ill-fated EV1 was charged using an inductive paddle, instead of actually plugging it into an outlet. In the ancient port area of Genoa, Italy, you’ll find electric buses that charge themselves wirelessly at the whopping rate of 60 kilowatts for 10 minutes each hour, by parking over flat charging coils built into the road surface. The system was built by Conductix-Wampfler of Weil am Rhein, Germany, which has also licensed patents from the University of Auckland.

These systems have long ago proved themselves able to move power wirelessly, often a lot of it, and with good efficiency. That’s possible because they transfer the power for only a short stretch through the air—a few tens of centimeters at most—nothing like the distances WiTricity and Intel are shooting for. The key question is whether engineering improvements will make greater separations practical.

“I’m skeptical about sending [power] over distances that are larger than the coil diameter,” says Menno Treffers, who works for Royal Philips Electronics and serves as the head of the two-year-old Wireless Power Consortium, which is aimed at establishing industry standards for the wireless charging of consumer electronics. Right now you can get wireless charging pads if you buy a Dell Latitude Z laptop or a Palm Pre smartphone. BlackBerry and iPhone owners can get this feature too, if they purchase special aftermarket charging sleeves.

The idea the Wireless Power Consortium is pushing is that eventually you’ll be able to buy a single charging pad that will recharge whatever device you place on top, regardless of brand. Treffers is keen to bring such interoperability to what he sees as a blossoming consumer technology, but he doesn’t expect it to get to the point where you can recharge your mobile gizmo while using it. “It’s not like you can charge your BlackBerry while sitting on the couch,” he says.

Eberhard Waffenschmidt, a Philips electrical engineer working with Treffers, has examined the question of what distances are possible for resonant inductive charging. His calculations suggest that the prototype systems that Intel and WiTricity have demonstrated are pushing the limits of what can practically be done without efficiency plummeting to ridiculously low levels. And even if poor efficiencies could be tolerated, the RF field levels required to send truly useful amounts of power over even modest distances would be above what you could reasonably expose people to. “All the journalists had missed this,” says Waffenschmidt, adding that “[charging] pads don’t have this problem.”

Is there no way then to increase the distance you can send power wirelessly? Of course there is, but not inductively. If you have a clear path, you could use microwaves or laser beams, as has been demonstrated many times. Or just keep it simple. “Sunlight is excellent for long-distance power transfer,” quips Treffers.

Even if resonant induction ends up being limited to short distances, it may yet have a great influence, particularly for the future of transportation. “We now [have the means to] charge a car safely and efficiently over gaps of 20 to 40 centimeters, and we believe we can build that into a roadway system,” says Covic. “That’s probably a decade away, but you’ve got to have a vision, and ours is roadway-powered systems.”

A small step in that direction is taking place at Berlin-based Bombardier Transportation, which is gearing up to offer an electric streetcar that is inductively powered through the roadway. Although there are other ways to avoid a streetcar’s hard-to-maintain cables [see “Fuel Cells Could Power a Streetcar Revival,” IEEE Spectrum, September 2009], Bombardier’s system, called Primove, avoids many of the problems that some of the competing solutions face.

If this approach for powering streetcars catches on, perhaps wider-ranging electric buses will be next. It’s not unreasonable to think that some decades from now private electric cars may also be able to draw at least some of their power wirelessly from suitably equipped roadways. Who knows? Maybe the early adopters of today’s electric cars will be able to retrofit their rides to get such a boost while cruising. Then even the people who bought their Tesla Roadsters in the 2010s could zoom along inductively powered highways of the future. Such a thing would surely please that roadster’s visionary namesake.