Ninety three million miles away and you can still feel its heat. We're talking about the sun, of course. It has a core temperature of 15 million degrees Celsius (27 million degrees Fahrenheit) and has been burning for nearly 5 billion years. Barring some unforeseen disaster, it should continue to burn for another 5 billion years.
The energy that lights up the sun and every other star in the universe comes from fusion -- the melding together of lighter atomic nuclei to form heavier nuclei. If fusion could be harnessed to generate electrical power here on earth, it would last us forever and would not contribute to the greenhouse effect, acid rain, or the depletion of the ozone layer. It would also be relatively safe. Unlike fission, where atoms are split rather than fused together, fusion can be done so as not to produce high-level radioactive by-products that must be carefully sealed in deep burial storage. Also, because fusion fuel is in the form of a plasma (an ionized gas), there is no solid fuel core that could melt down -- the so-called "China Syndrome."
In July of 1976, a two-week meeting was held at the Claremont Hotel in Berkeley, California, to discuss an idea for using accelerator technology to produce fusion energy that was first proposed by Alfred W. Maschke, a physicist who was then at Brookhaven National Laboratory. This meeting, which was conducted jointly by the Lawrence Berkeley Laboratory and the Lawrence Livermore National Laboratory, brought together some 50 researchers from LBL, LLNL, and other national labs. The issue they discussed was whether intense, short bursts of high powered heavy ion beams could ignite thermonuclear fuel confined by its own inertia so as to produce a net gain of energy.
The consensus was that the idea merited serious attention. A number of small preliminary studies were launched including several at LBL's Accelerator and Fusion Research Division (AFRD), where scientists and engineers began working on ways of accelerating, transporting, and focusing heavy ion beams at higher currents than ever before. The success of this early research was recognized in 1983 when the U.S. Department of Energy (DOE) established a Heavy Ion Fusion Accelerator Research (HIFAR) program. Today, the HIFAR group at LBL, in collaboration with a similar group at LLNL, is poised on the brink of another milestone -- a project designed to answer some of the crucial questions that will help decide whether fusion becomes the energy source of the next century or remains an elusive dream. The project is called ILSE.
The great lure of fusion is twofold. First, a minuscule amount of fuel translates into a prodigious output of energy. When nuclei are fused together, the mass of the new nuclei is less than the total of the originals. This extra mass is converted into energy in accordance with Einstein's equation, E=mc2. The most easily attained fusion reaction involves fusing nuclei of the two isotopes of hydrogen, deuterium and tritium, to make nuclei of helium. The fusion of a single milligram of deuterium and tritium produces the same energy released in the explosion of 80 kilograms (about 175 pounds) of TNT.
The second factor behind the lure of fusion is that the supply of these hydrogen isotopes is almost infinitely abundant. Deuterium, for example, comes directly from water. There is enough of it that could be extracted from the top 10 inches of Lake Superior to supply the electricity needs of the United States at present consumption levels for approximately 5,000 years. Tritium is produced from lithium which can also be obtained from water.
High temperatures and density are the conditions needed for fusion to take place. Since nuclei carry positive electrical charges, they normally repel one another. Heat causes the nuclei to move, however, and if the temperatures are sufficiently high, the nuclei will collide with enough speed to overcome the force of repulsion. These conditions are created in the sun and other stars by immense gravitational pressure, but the earth's gravity is too weak to provide the necessary density. To achieve fusion on earth, fusionable atoms must be heated to temperatures of 100 million degrees Celsius (180 million degrees Fahrenheit) and confined long enough for the heated nuclei to interact.
There are two alternatives to gravitational confinement being pursued for earthbound fusion. Of the two, the older and perhaps more widely familiar is "magnetic confinement," also known as magnetic fusion energy (MFE). In the leading MFE approach, thermonuclear fuel is contained as a plasma within a doughnut-shaped reactor (called a tokamak) by a powerful magnetic field and heated to ignition. Although experimental efforts have shown tantalizing promise, there remain many technical obstacles in scaling the physics up from small laboratory experiments to commercial-size reactors. So far, these experiments have consumed more energy than they produced.
The other alternative to gravitational confinement is "inertial confinement," also known as inertial fusion energy (IFE). With IFE, a hollow shell of frozen deuterium and tritium is set inside a sphere about the size of a pea. This frozen shell of thermonuclear fuel is surrounded by an outer "ablation" layer that is rapidly heated until it turns to plasma. The plasma flying out from the ablation layer implodes the fuel, compressing its density about a thousand times and causing it to burn. If the fuel burns rapidly enough, it is confined by its own inertia and requires no external confinement system.
In contrast to magnetic confinement, the ability of an inertial confinement system to produce far more energy than it uses has been thoroughly demonstrated. Unfortunately, the proof- of-concept model is a hydrogen bomb. Thus for IFE, the issue is whether the technology can be scaled-down to useful size.
The technical problems in IFE can be separated into three categories -- the design of a "driver" that can rapidly heat the ablator, the design of a target of thermonuclear fuel that can be imploded into ignition, and the design of a reactor that can convert the energy released by the ignited fuel into electricity. ILSE is designed to model the beam manipulations required of a driver.
It takes about 75 watts of power to light a room sufficiently for reading. It takes 100 to 1,000 trillion watts of power to ignite a target of thermonuclear fuel. This kind of power must be attained almost instantaneously and must be delivered in pulses of about 10 billionths of a second (10 nanoseconds) in duration. Experimental targets have been successfully heated (but not to ignition) using lasers. These studies have provided valuable knowledge about target physics. For commercial IFE applications, however, a driver must be able to deliver several shots of energy a second (analogous to the repeated mini-explosions in the cylinder of an automobile engine). Also, to be cost-effective, the energy delivered to the target must be at least 10 percent of the energy energy required to power the driver. In addition to this high rate of repetition and efficiency, a commercial driver will also need good reliability (better than 90 percent) and a lifetime of about 30 years. At this time, lasers cannot meet these requirements.
Based on a number of studies conducted during the 1980's, including two from the National Academy of Sciences, DOE has concluded that the most promising driver for a commercial-scale reactor would be high-powered beams of heavy ions, such as xenon, mercury, or lead. Such ions deposit their energy at high voltages over a short distance, which means that incoming beams can be set up to produce a very high energy density in a target. This is a critical requirement for igniting the fuel. Furthermore, much of the technology pertaining to the particle accelerators that would create and deliver these beams has already been developed for the giant machines that are now used in high-energy physics research. However, there are a number of problems that must be resolved.
High beam power is traditionally achieved with high currents. Obtaining adequate power requires total currents greater than 10,000 amperes. Induction linacs, linear accelerators that induce an electromotive force on ions by rapidly changing the strength of a magnetic field inside a vacuum cavity, have proven capable of producing the necessary current. Transporting and focusing such intense beams is another story. Problems arise as a result of "space-charge" forces -- the mutually repulsive forces between so many positively charged ions.
By 1985, LBL's HIFAR group with their Single Beam Transport Experiment had demonstrated that heavy ion beams could be transported and focused at currents several times higher than was first believed possible in beams dominated by space-charge forces. These currents, however, were still far short of what was needed. Under the leadership of physicist Denis Keefe, the group switched from working with a single high current beam to accelerating and transporting a number of independently focused, less intense beams. The idea was that the combined energies of these multiple beams, when overlapped on a target, would be much easier to control and could even be more effective than a single high energy beam. This work culminated in MBE-4, the world's first induction linac capable of accelerating and focusing four parallel beams simultaneously. The MBE-4 experiments, which were recently completed, demonstrated that it is possible to amplify the currents of multiply focused heavy ion beams (through the manipulation of their voltages) during acceleration without sacrificing beam quality or control.
MBE-4's beams of cesium ions were each accelerated to nearly one million electron volts (1 MeV) of kinetic energy and amplified to currents of up to 90 milliamps per beam. This was impressive by the standards of what had been done in the past, but still quite small by the standards of a commercial-scale driver. DOE's National Energy Strategy calls for a demonstration fusion reactor (using either IFE or MFE) by the year 2025, and a commercial power plant by about 2040. For these goals to be met, scientific and technical questions about controlling and manipulating multiple heavy ion beams on the scale of a commercial driver must be answered. That is the the mission of ILSE.
ILSE stands for Induction Linac Systems Experiments. The project's main feature is a 50 meter long linear accelerator designed to provide four beams of heavy ions at the same current and space-charge density as would be used in a full-size multiple beam IFE driver.
Roger Bangerter is a physicist who has been advocating the use of heavy ion beams as an IFE driver since the concept was proposed. While at LLNL, he served as co-chair of that first meeting in the summer of 1976 (along with David Judd, of LBL). He went on to become the deputy division leader of that laboratory's X Division, which oversees IFE theoretical work, and also headed up LLNL's Inertial Confinement Fusion Application Group, which designs fusion reactors. Following the death of Denis Keefe in 1990, Bangerter became the leader of LBL's HIFAR group.
In discussing the importance of ILSE to a DOE review committee, he stated: "ILSE will address many of the remaining scientific and technical issues at a relevant scale. It will be a flexible experimental tool that should provide the data needed to determine the feasibility and cost of heavy ion fusion."
The DOE review committee subsequently endorsed ILSE with excellent marks.
LBL physicist Thomas Fessenden is the project leader for ILSE. He says, "We see ILSE as a science project, a model experiment in accelerator physics that will enable us to test the beam manipulations required of a heavy ion fusion driver."
ILSE starts with an ion source and injector that can generate four beams of heavy ions, probably neon or potassium, at energies of about 2 MeV and a full ampere of current. The temperature of each beam coming out of the injector is very low - - approximately one tenth of an electron volt -- so that the ions have very little motion relative to each other.
This is critical as Bangerter explains. "For a commercial- scale driver, accurate focusing is possible only if the beams are heated no more than four orders of magnitude during acceleration. The final temperature needs to be less than one keV (one thousand electron volts), otherwise there is too much random motion in the ions to obtain the necessary small focus."
The beams that emerge from the injector are round in shape. Prior to acceleration, they must pass through a "matching section" where their round shape is molded into an "alternating gradient" profile that matches the focusing fields that transport the beams through the accelerator. Fessenden likens this to "taking a soda straw and pinching it first one way, then another." The beams are also "squeezed together" to further facilitate their acceleration.
From the matching section, the beams are ushered through two acceleration sections. In the first and smaller of these sections, the focusing force arises from electrical fields generated by a series of electrostatic quadrupoles. Once the four beams have been accelerated to 4.5 MeV, they are combined into a single beam and matched into a longer second acceleration section which focuses them through a series of magnetic quadrupoles. Magnetic focusing will be required for the highly energetic and intense beams of heavy ions used in an actual IFE driver. The choice of mid-sized ions like neon and potassium for ILSE provides useful data at a fraction of the cost of heavier ions by permitting the study of magnetic focusing at energies of only 5 to 10 MeV.
Initially, only one of the four electrostatically-focused beams will be ferried into the magnetically-focused section for acceleration to the full 10 MeV. A single beam is all that is needed to test the linac's ability to produce, accelerate, and focus the ions. The linac tests will also examine tradeoffs between actually steering the beam or simply aligning it through the focusing system, and will look at various techniques for controlling the shape of the beam which is important for achieving high energy gain at the target.
The combining of the multiple beams will take place as a part of ILSE's experimental program. This phase of the project will ultimately entail the adding on of beam transportation sections that will give ILSE a J-shaped configuration. The bend is needed to determine how well space-charge dominated beams of heavy ions can be transported through large angles. Almost every IFE scheme calls for the target to be heated from more than one side. This requires bending some of the multiple beams of heavy ions through angles as large as 270 degrees. While bending magnets are routinely used to handle heavy ion beams in particle accelerators, these beams have always been at low currents.
Another critical beam manipulation to be studied in ILSE's experimental program involves an "energy tilt" at the exit of the accelerator. This phenomenon of beam physics, in which the head of the beam moves slower than its tail, shortens the beam and increases its current, thereby boosting the power of the beam. However, an energy tilt also makes the beam impossible to focus. ILSE will feature a drift-compression and power amplification section that will boost the power of the beam while removing an energy tilt of more than ten percent prior to its entering a final focusing section.
Although ILSE will not be used to ignite any thermonuclear fuel, it will give scientists at LBL their first opportunity to focus a heavy ion beam onto a target. Different schemes for directing the beams into a target will be explored. One of these schemes might be the "z-pinch transport," in which a large electrical field applied inside a magnetic field pinches the beam into a narrow focus. ILSE's beams will also give scientists their first chance to model the behavior of high current heavy ion beams in the partial vacuum of a target chamber.
If all goes well in ILSE's linac and experimental programs, the ends of the J could eventually be connected to form a complete ring, making ILSE the first high current heavy ion "recirculator" -- a fast-cycling circular induction accelerator. A recirculator is like a synchrotron in that the ion beam is sent around through the acceleration sections a multiple number of times to further increase its energy. For example, a recirculator configuration might boost ILSE's total beam energy to 100 MeV.
Conceptual studies have shown that recirculators could reduce the cost of a heavy ion driver. However, says Fessenden, "No one knows exactly how many times beams of such high currents can be recirculated. Inserting the beam into the recirculator and extracting it back out are two of the key manipulations that must be studied."
Current plans call for ILSE to be located adjacent to LBL's storied Bevatron accelerator. Construction would start in the fall of 1993 and be completed in 1998. Full operation and experimentation would commence sometime in 1999. The total projected cost of the project is approximately 70 million dollars.
If the physics and engineering studies at ILSE are favorable, the next step would be the construction of an intermediate-sized driver facility around the year 2005. This facility would advance IFE technology closer to that needed for a commercial driver and would enable researchers to begin target physics studies with a heavy ion beam, comparable to the studies that have been done with lasers. This would put IFE on schedule to meet the goals of the National Energy Strategy.
In addition to Bangerter and Fessenden, other key LBL staff members in the ILSE project include Craig Fong, the project manager, Edward Lee and Andy Faltens, the project advisors, Lou Reginato, the project electrical engineer, Craig Peters, the project mechanical engineer, Henry Rutkowski, who is designing ILSE's ion source, and Greg Raymond, the project architect.