The Dark-Horse Lab That Just Might Figure Out Fusion

Discover Magazine By Tim Folger | January 26, 2011 | Republished here Nov. 30, 2015

It is the energy source that could change the world. It has eluded every effort to master it. But Glen Wurden thinks he knows how to tame the heart of the sun.

The Shiva Star facility, located at the Kirtland Air Force Base in Albequerque, produces a power pulse that justifies its name.
Air Force Research Laboratory

Everything was ready. The lab’s blast walls were up; sheets of Kevlar hung to catch shrapnel; banks of capacitors on the co-opted Air Force experiment primed to unleash 12 million amps of current; X-ray detectors set to snag bursts of photons; bottles of celebratory champagne chilled. If all went as planned, Glen Wurden would be a very happy man, and one experiment closer in his long-shot effort to exploit a nearly limitless source of energy.

The champagne had to wait, though, because something went wrong. Not a glitch, not a minor mishap. An explosion. “The floor shook; the walls shook; there was a hell of a boom,” says Wurden, a sandy-haired, 54-year-old physicist and fusion program manager at Los Alamos National Laboratory in New Mexico. “An Air Force guy with us said, ‘Damn, that was loud!’ One of the Kevlar blankets was tossed 40 feet. A piece of shrapnel went through one of the air-conditioning ducts. The experiment was a spectacular failure.”

For someone talking about a fantastically borked bit of work, Wurden sounds almost cheerful. We are sitting in his office at Los Alamos, a 45-minute drive northwest of Santa Fe, on a cool September afternoon, nine months after the explosion. No one was hurt that day—the blast walls protected Wurden and his colleagues from the, uh, mistake in the adjoining room. “It turns out there was a short circuit,” Wurden says as we look on his computer monitor at photos of a blackened, debris-strewn lab. The short overloaded the equipment with 18 million amps, delaying for a few months the initial test of technology that could yield the world’s first commercially viable nuclear fusion reactor.

“My goal in life is to make fusion energy happen. Period,” Wurden says. The control of nuclear fusion—the reaction that powers stars and hydrogen bombs—would permanently solve the world’s energy problems, not to mention a few geopolitical ones. No small ambition, by any measure. But Wurden harbors another goal, nearly as daunting. He wants to beat the world’s two biggest fusion projects in the race to make fusion not just possible but practical. The competition could scarcely be more lopsided. The ranking fusion heavyweights, one in France and one in California, each have at least a thousand-to-one funding advantage over Wurden’s project and a huge edge in manpower as well. ITER—an international fusion experiment now under construction in the south of France—will probably cost $20 billion by the time it is finished in 2018. The $3.5 billion National Ignition Facility (NIF) in Livermore, California, is slated to begin fusion tests by 2012, after 15 years of construction and development.

When I tell Wurden that I would like to compare his research favorably with the work going on at NIF and ITER, he tries to discourage me. “I’m not comfortable with that. Why should we be in the same paragraph with a $4 billion machine? We’re not even a $4 million operation.” Yet after I spend a couple of afternoons with him, it becomes clear that he thinks he has a reasonable shot at something that has eluded researchers for decades: creating a fusion device that yields more energy than it consumes, and doing so on a budget of tens of millions instead of billions.

“Our goal is to demonstrate fusion gain before ITER is built. I don’t know about beating NIF—it will probably get a good shot in a year or two. We almost certainly won’t make that,” Wurden says. “Our goal is to make a fusing plasma in the lab once a week.” Without any more explosions.

The physics of fusion seems so straightforward, so tantalizingly graspable: Two atomic nuclei (the dense centers of atoms) collide and merge to form a larger nucleus, releasing a lot of energy. Fusion is the opposite of fission, which frees energy when an atom like uranium splits into two smaller atomic nuclei. Today’s nuclear power plants use the heat from uranium fission reactions to do nothing more complicated than boil water, making pressurized steam that spins turbines to generate electricity. Fusion power plants would also be boilers, albeit exceedingly complex ones. Although physicists and engineers conceptually figured out how to construct fission power plants almost immediately after the invention of the atomic bomb, fusion has never been tamed.

The quest to control the process has seduced physicists for more than half a century, for obvious reasons. As an energy source, it seems too good to be true. It releases no greenhouse gases. Its primary fuel—deuterium, a heavy version of hydrogen—can be extracted from seawater. Unlike the waste from conventional nuclear power plants, which remains radioactive for tens of thousands of years, the by-products of fusion decay within decades. And there is irrefutable evidence that stable, self-sustaining fusion reactions are physically possible. The most obvious example is a fusion reactor that has been running smoothly for more than 4 billion years: the sun, which fuses hundreds of millions of tons of hydrogen nuclei into helium every second. Without fusion, stars would not shine. Just look up at the sky and you can see that fusion reactors are stupefyingly common.

The huge National Ignition Facility aims to achieve fusion reactions using the world’s most powerful lasers.
Jacqueline McBride/ LLNL

So why, on Earth, is fusion so difficult? The technical issues are formidable, to be sure, starting with a fuel so hot that it would vaporize any known material, and positively charged nuclei that want to repel each other rather than fuse together. The sun and other stars create the extraordinary conditions necessary for fusion—temperatures of about 15 million degrees Celsius and pressures 200 billion times greater than Earth’s atmospheric pressure—by dint of sheer gravity. A star must have at least a tenth of the sun’s mass in order to ignite. The universe is littered with failed fusion reactors, so-called brown dwarfs, which never scooped up enough matter to get their fusion reactors going.

Despite those astronomical obstacles, Wurden says the biggest impediment to controlled fusion has been a lack of resources. “If you gave me an infinite amount of money, I couldn’t produce a fusion reactor tomorrow. On the other hand, if you gave me a billion dollars a year for 20 years, that’s the other extreme. Is there a way to do it faster? Yeah, I think there is. Somewhere between one day and 20 years is the right number.”

There is an old joke among fusion scientists: “Fusion energy is 20 years away and always will be.” Utopian fusion forecasts have a long history. At a United Nations atomic energy conference held in Geneva in 1955, an Indian nuclear physicist named Homi Bhabha said, “I venture to predict that a method will be found for liberating fusion energy in a controlled manner within the next two decades. When that happens the energy problems of the world will truly have been solved forever.”

So has anything changed? Will fusion energy still be as distant a dream three decades from now as it was three decades ago? If Wurden believed that, he would be in another line of work. But here he is, searching for some way to finesse the process, accomplishing by subtler means what nature does by brute force. Manipulating sun-size volumes of gas and titanic gravitational forces are not viable options. Wurden’s strategy? Call it fusion in a beer can.

The technical term for what Wurden and his colleagues at Los Alamos do is “magnetized target fusion.” One component of their experimental device is a hollow aluminum cylinder, 30 centimeters tall and 10 centimeters in diameter, with walls 1 millimeter thick. “It’s the size of an extra-big beer can—Australian size,” Wurden tells me as we approach his lab, which occupies all of a hangar-size building next door to his office. On the way into the lab I notice mouse droppings on the floor and some discarded, damaged equipment on shelves with tags labeled “NFG” attached. “That means no good,” Wurden says, omitting an adverb. “Kind of like FUBAR, you know?”

Banks of refrigerator-size capacitors, devices for storing and releasing electrical energy, take up most of the lab’s floor space, all linked by thick colored cables. But it is the can that will be the center of the action. That is where the fusion will happen—assuming the absence of short circuits and explosions. With those caveats, canned fusion is supposed to work like this: 1. Switch on powerful magnetic fields to create and control a hot plasma (ionized gas) of deuterium. 2. Trap the plasma inside the aluminum can, again using magnetic fields; if the plasma physically touched the can, the aluminum would vaporize almost instantly. 3. Apply a quick burst of current to crush the can in a few millionths of a second, heating the confined deuterium to millions of degrees and compressing it to such an extreme that fusion occurs. 4. Harness the released energy and change the world.

“We crush the can in 24 microseconds with 12 million amps of current,” Wurden says. “The magnetic fields are held inside and go up by a factor of 100. There are no magnetic fields like that in our solar system, not even in this corner of the galaxy. The pressure inside equals a million atmospheres; the temperature goes up by a factor of 30 to 100.”

Last April, Wurden finally combined all the elements of canned fusion into a single device and put it through an engineering shakedown. Here at Los Alamos, he and colleagues are now working on better ways to create and manipulate the deuterium plasma. The can-crushing experiments—and the odd explosion—are taking place at Kirtland Air Force Base in Albuquerque.

Wurden’s plasma-holding apparatus is a quartz tube, 40 centimeters in diameter and 150 centimeters long, mounted on a table and partially wrapped in the aluminum coils of an electromagnet. Pumps for evacuating the tube and diagnostic equipment for studying the plasma surround the table. Three computers control it all from a small room at the other end of the lab, which is sealed off by blast doors whenever the experiment runs.

The computer-controlled electromagnets are designed to generate potent electromagnetic fields inside the quartz tube. Those fields are supposed to ionize the deuterium and form it into a football-shaped plasma, a geometry that would prevent it from squirting out of the collapsing can. Maintaining that exact geometry is crucial to the experiment’s success, because leaks would rob the plasma of the pressure and energy needed for fusion. Wurden’s efforts now focus on carefully measuring the plasma in the quartz tube to check for the right density, temperature, and stability to make fusion possible.

“The diagnostics all happen in about 60 microseconds,” Wurden says. “The signals are sampled very quickly and stored on the computer so we can understand what was happening in the plasma. There are lots of ways to fail. We’re merging this can-compression technology with plasma-formation technology that we’ve studied for 20 years into one entity. Our first test shot of the can crusher was December 19, 2008.” That was the short-circuited, explosive experiment. “The can didn’t collapse very much because the current didn’t get to it. We were kind of depressed.” But a second shot a few months later, again with an empty can, went as planned, setting the stage for last April’s full can-and-plasma test.

The machine that crushes the cans began life as a weapon, according to some reports, developed at Kirtland Air Force Base to destroy incoming ballistic missiles with bursts of high-energy plasma. It is called Shiva Star, after the Hindu god of destruction, and for the starlike pattern formed by six rows of capacitor banks arrayed radially around a two-meter-high hub of electrodes and coils. The entire setup measures 25 meters in diameter and fills a room in a nondescript three-story building on the air force base. Jim Degnan, a physicist at Kirtland who collaborates with Wurden’s team, insists that Shiva Star has never been used as a weapon, but he has seen it destroy a fair number of experiments.

“We operate at 5 megajoules,” he says. “That’s like two and a half pounds of high explosives. It can break vacuum vessels, throw shrapnel. Occasionally we start a fire. The fire department tells us if we can put it out with one extinguisher, we don’t have to evacuate the building.”

For the crushing tests, the can is mounted vertically under Shiva Star’s central column, which is stabilized with four tons of weights to prevent magnetic fields from pulling the structure apart. Shiva Star discharges 12 million amps in 10 microseconds. The can’s collapse must be nearly instantaneous and perfectly symmetrical to contain the deuterium plasma long enough to spark fusion. “In this experiment, the technical difficulty is getting all the different capacitor discharges to work at the right times with the right currents,” Degnan says, “and getting all the diagnostic equipment to work, because if you can’t measure it, you don’t know what happened.”

For Wurden, April’s Shiva Star experiment with the plasma-filled can prompted a surge of optimism. In two or three years, he and Degnan should have a rough idea of whether anything insurmountable stands between them and a fusion reactor that can actually produce a net surplus of energy.

“If it works the way we hope it will, and if we do enough tests to find what makes it better or worse, we’ll be ready to go to a bigger facility,” Wurden says. “If we go up four or five times in size and switch to a fuel with a mix of deuterium and tritium,” another type of heavy hydrogen, “we should have break-even plasma conditions—if things work out.”

This plasma injector tube at Los Alamos allows Glen Wurden to fine-tune his fusion experiment.
Courtesy of Los Alamos National Laboratory

“Break-even” is the goal of all fusion researchers. This is the point at which the amount of energy coming out of the fusion reactions equals the amount of energy needed to maintain the plasma. No fusion experiment has yet reached that point. More distant still is a reactor able to generate substantially more energy than it consumes. An economically viable fusion power plant would probably have to do 30 times better than break-even.

Wurden’s two big rivals have a reasonable shot at reaching or surpassing break-even within a decade. ITER, funded by the United States, the European Union, Russia, China, India, Japan, and South Korea, is scheduled to run its first fusion experiments in 2018 or 2019. It will use the most advanced in a line of fusion machines called TOKAMAKS, an acronym for the Russian phrase meaning “toroidal chamber with magnetic coils.” Many physicists see TOKAMAKS as the most promising path to fusion energy. Inside ITER’s enormous, doughnut-shaped reactor walls, magnetic fields, electric currents, microwaves, and particle beams will heat a deuterium-tritium plasma to fusion temperatures for about 20 minutes. During that time, ITER’s designers anticipate that the reactor will put out 500 megawatts of power while using only 50 megawatts, a tenfold energy gain.

The National Ignition Facility at Lawrence Livermore National Laboratory, located 45 miles east of San Francisco, houses the world’s most powerful lasers, 192 of them. Sometime within the next year or two, for a few billionths of a second, those beams will focus 500 trillion watts on a pellet of deuterium and tritium housed inside a little capsule about the size of an Advil tablet. If all works as expected, the deuterium and tritium will slam together at about a million miles per hour, forming a plasma, fusing, and perhaps producing a net energy gain.

Wurden thinks his team has a shot at beating ITER to the break-even finish line, but only if he can scrounge up a little more cash. “We can’t do it with the funding we have now,” he says. “The Department of Energy sponsors all the magnetic fusion research in the country. Alternate projects like ours are at best about 10 percent of the budget, maybe $20 million divided among 10 universities and a couple of national labs.”

Despite the way the government is placing its bets, canned fusion has potential advantages over both of the big projects. It falls somewhere between NIF and ITER in its strategy. Wurden expects his experiment to create a plasma higher in density than ITER’s but lower than NIF’s compressed-pellet plasma. Dense plasmas favor more fusion reactions; so do longer confinement times—but sun-hot plasmas are exceedingly difficult to control. Wurden will trap his plasma for a few millionths of a second. ITER, with its less dense plasma, will need to confine it for full seconds at a time. NIF requires just billionths of a second, but at the cost of enormous energy output from its high-power lasers, which can fire only two or three times a day because they need many hours to cool down.

Canned fusion, if it works, would need less power than NIF and shorter plasma-confinement times than ITER; as a result, a working reactor based on its principles might cost tens of millions of dollars rather than billions. If fusion is ever going to make commercial sense, that is exactly the kind of economic breakthrough it will require.

Even if NIF beats canned fusion to break-even, Wurden thinks his approach will be more practical in the long run. NIF’s lasers currently fire just two or three times a day. It takes 30 minutes just to position the fuel capsule. A commercial laser fusion reactor might have to fire about 15 times a second.

“There’s another problem,” Wurden says. “NIF’s targets are cryogenic”—the fuel pellets are frozen to –255 degrees Celsius—“and in a power plant your poor little frozen target would have to fly into this nuclear hell and not get burned up on the way in. It’s a big problem. There are a lot of questions as to whether you could do this at any cost. With magnetized target fusion, we talk about firing once every 10 seconds. We have a lot of problems too, but we’re not talking about 15 times a second. There are a lot of reality checks that need to be done for anybody’s reactor proposal.”

For a reality check on Shiva Star, I spoke with Jaeyoung Park, an experimental physicist who has taken leave from Los Alamos to join a small team in Santa Fe that is pursuing its own fusion research. His biggest concern is that Wurden may not be able to contain the deuterium plasma long enough. “It’s very difficult to squeeze the plasma uniformly—the squeezing has to be fast and furious,” Park says. “And heat losses might make it impossible for the plasma to achieve the high temperatures needed for fusion. But Glen is planning some significant experiments, and even if the first ones fail, the results should tell us something important.”

Wurden acknowledges those problems and brings up another for good measure. “How do you control millions of amps of current at thousands of volts?” he asks. “The switches we use are fancy things that work under high voltage. We can switch high currents maybe 20 times a day.” But a working fusion reactor based on Shiva Star would need to handle such currents once every 10 seconds. He is on a roll now, plotting the path that could finally bring fusion energy to your household electrical outlet: “Maybe years from now, some of these things that are difficult to do today will be technologies we can handle. We’re at the Wright brothers stage now. If someone had given the Wright brothers the plans for a 747 and told them to just build it, well, they wouldn’t know what a jet engine is, and certainly wouldn’t have the ability to make turbine blades or alloys.

“If someone tells you we’ll have fusion energy in three years, that is not going to happen. Even if you want to solve the energy problem in the next 30 years, fusion is not the answer, and I say that as a fusion scientist. If you want to solve it in the 50-to-100-year time­scale—yeah, I think it is the answer. I like to ask other physicists, ‘How many miracles do you need for your concept to work?’ There isn’t a single concept that doesn’t need an engineering miracle or two or three. Not one. But if you can count the number of miracles on one hand, you might say your concept is viable—you only need one handful of miracles. That’s where fusion research is: How many miracles do you need?”


Fusion energy research is the ultimate high-risk, high-reward scientific endeavor. So who better to tackle the challenge than entrepreneurs? “ITER and NIF have a heck of a lot of money, but they also have incredibly complex technology,” says Doug Richardson, CEO of British Columbia–based General Fusion. “We looked at the endgame from the very beginning: Can you put this into a power plant easily?”

To that end, Richardson and his competitors are scheming up novel reactor designs. General Fusion’s magnetized target fusion reactor will incorporate a multipurpose liquid-metal lining to produce tritium, protect equipment from damage, and extract the heat that generates energy. The company hopes to achieve break-even by 2013. Two other companies, Energy Matter Conversion Corp. (EMC2) and Tri Alpha Energy, are developing reactors that use proton-boron fuel, which requires even higher temperatures than deuterium does but allows almost direct conversion of fusion into electricity, without boiling water to drive a generator.

Still, it’s not easy fiddling with million-degree plasmas on a tight budget. With the exception of Tri Alpha, which secured $40 million in private funding (some of it from Microsoft cofounder Paul Allen) three years ago, fusion has proved too risky for even the most daring venture capitalists. General Fusion limped through its first four years supported by a couple hundred thousand dollars from family and friends. “Skepticism and credibility are our top challenges by far,” Richardson says. Andrew Grant