Inertial confinement fusion (ICF) is a process where nuclear fusion reactions are initiated by heating and compressing a fuel target, typically in the form of a pellet that most often contains a mixture of deuterium and tritium.
To compress and heat the fuel, energy is delivered to the outer layer of the target using high-energy beams of laser light, electrons or ions, although for a variety of reasons, almost all ICF devices to date have used lasers. The heated outer layer explodes outward, producing a reaction force against the remainder of the target, accelerating it inwards, and sending shock waves into the center. A sufficiently powerful set of shock waves can compress and heat the fuel at the center so much that fusion reactions occur. The energy released by these reactions will then heat the surrounding fuel, which may also begin to undergo fusion. The aim of ICF is to produce a condition known as "ignition", where this heating process causes a chain reaction that burns a significant portion of the fuel. Typical fuel pellets are about the size of a pinhead and contain around 10 milligrams of fuel: in practice, only a small proportion of this fuel will undergo fusion, but if all this fuel were consumed it would release the energy equivalent to burning a barrel of oil.
ICF is one of two major branches of fusion energy research, the other being magnetic confinement fusion. To date most of the work in ICF has been carried out in the United States, and generally has seen less development effort than magnetic approaches. When it was first proposed, ICF appeared to be a practical approach to fusion power production, but experiments during the 1970s and 80's demonstrated that the efficiency of these devices was much lower than expected. For much of the 1980s and 90s ICF experiments focused primarily on nuclear weapons research. More recent advances suggest that major gains in performance are possible, once again making ICF attractive for commercial power generation. A number of new experiments are underway or being planned to test this new "fast ignition" approach.
Fusion reactions combine lighter atoms, such as hydrogen, together to form larger ones. Generally the reactions take place at such high temperatures that the atoms have been ionized, their electrons stripped off by the heat; thus, fusion is typically described in terms of "nuclei" instead of "atoms".
Nuclei are positively charged, and thus repel each other due to the electrostatic force. Counteracting this is the strong force which pulls nucleons together, but only at very short ranges. Thus a fluid of nuclei will generally not undergo fusion, the nuclei must be forced together before the strong force can pull them together into stable collections. Fusion reactions on a scale useful for energy production require a very large amount of energy to initiate in order to overcome the so-called Coulomb barrier or fusion barrier energy. Generally less energy will be needed to cause lighter nuclei to fuse, as they have less charge and thus a lower barrier energy, and when they do fuse, more energy will be released. As the mass of the nuclei increase, there is a point where the reaction no longer gives off net energy — the energy needed to overcome the energy barrier is greater than the energy released in the resulting fusion reaction.
The key to practical fusion power is to select a fuel that requires the minimum amount of energy to start, that is, the lowest barrier energy. The best fuel from this standpoint is a one to one mix of deuterium and tritium; both are heavy isotopes of hydrogen. The D-T (deuterium & tritium) mix has a low barrier because of its high ratio of neutrons to protons. The presence of neutral neutrons in the nuclei helps pull them together via the strong force; while the presence of positively charged protons pushes the nuclei apart via Coloumbic forces (the electromagnetic force). Tritium has one of the highest ratios of neutrons to protons of any stable or moderately unstable nuclide -- two neutrons and one proton. Adding protons or removing neutrons increases the energy barrier.
In order to create the required conditions, the fuel must be heated to tens of millions of degrees, and/or compressed to immense pressures. The temperature and pressure required for any particular fuel to fuse is known as the Lawson criterion. These conditions have been known since the 1950s when the first H-bombs were built.
In a "hydrogen bomb" the fusion fuel is compressed and heated with a separate fission bomb. A variety of mechanisms transfers the energy of the "trigger"'s explosion into the fusion fuel. The use of a nuclear bomb to ignite a fusion reaction makes the concept less than useful as a power source. Not only would the triggers be prohibitively expensive to produce, but there is a minimum size that such a bomb can be built, defined roughly by the critical mass of the plutonium fuel used. Generally it seems difficult to build nuclear devices smaller than about 1 kiloton in size, which would make it a difficult engineering problem to extract power from the resulting explosions. Also the smaller a thermonuclear bomb is, the "dirtier" it is, that is to say, the percentage of energy produced in the explosion by fusion is decreased while the percent produced by fission reactions tends toward unity (100%). This did not stop efforts to design such a system however, leading to the PACER concept.
If some source of compression could be found, other than a nuclear bomb, then the size of the reaction could be scaled down. This idea has been of intense interest to both the bomb-making and fusion energy communities. It was not until the 1970s that a potential solution appeared in the form of very large, very high power, high energy lasers, which were then being built for weapons and other research. The D-T mix in such a system is known as a target, containing much less fuel than in a bomb design (often only micro or milligrams), and leading to a much smaller explosive force.
Generally ICF systems use a single laser, the driver, whose beam is split up into a number of beams which are subsequently individually amplified by a trillion times or more. These are sent into the reaction chamber (called a target chamber) by a number of mirrors, positioned in order to illuminate the target evenly over its whole surface. The heat applied by the driver causes the outer layer of the target to explode, just as the outer layers of an H-bomb's fuel cylinder does when illuminated by the X-rays of the nuclear device.
The material exploding off the surface causes the remaining material on the inside to be driven inwards with great force, eventually collapsing into a tiny near-spherical ball. In modern ICF devices the density of the resulting fuel mixture is as much as one-hundred times the density of lead, around 1000 g/cm³. This density is not high enough to create any useful rate of fusion on its own. However, during the collapse of the fuel, shock waves also form and travel into the center of the fuel at high speed. When they meet their counterparts moving in from the other sides of the fuel in the center, the density of that spot is raised much further.
Given the correct conditions, the fusion rate in the region highly compressed by the shock wave can give off significant amounts of highly energetic alpha particles. Due to the high density of the surrounding fuel, they move only a short distance before being "thermalized", losing their energy to the fuel as heat. This additional energy will cause additional fusion reactions in the heated fuel, giving off more high-energy particles. This process spreads outward from the center, leading to a kind of self sustaining burn known as ignition.
The primary problems with increasing ICF performance since the early experiments in the 1970s have been of energy delivery to the target, controlling symmetry of the imploding fuel, preventing premature heating of the fuel (before maximum density is achieved), preventing premature mixing of hot and cool fuel by hydrodynamic instabilities and the formation of a 'tight' shockwave convergence at the compressed fuel center.
In order to focus the shock wave on the center of the target, the target must be made with extremely high precision and sphericity with aberrations of no more than a few micrometres over its surface (inner and outer). Likewise the aiming of the laser beams must be extremely precise and the beams must arrive at the same time at all points on the target. Beam timing is a relatively simple issue though and is solved by using delay lines in the beams' optical path to achieve picosecond levels of timing accuracy. The other major problem plaguing the achievement of high symmetry and high temperatures/densities of the imploding target are so called "beam-beam" imbalance and beam anisotropy. These problems are, respectively, where the energy delivered by one beam may be higher or lower than other beams impinging on the target and of "hot spots" within a beam diameter hitting a target which induces uneven compression on the target surface, thereby forming Rayleigh–Taylor instabilities in the fuel, prematurely mixing it and reducing heating efficacy at the time of maximum compression.
All of these problems have been substantially mitigated to varying degrees in the past two decades of research by using various beam smoothing techniques and beam energy diagnostics to balance beam to beam energy though RT instability remains a major issue. Target design has also improved tremendously over the years. Modern cryogenic hydrogen ice targets tend to freeze a thin layer of deuterium just on the inside of a plastic sphere while irradiating it with a low power IR laser to smooth its inner surface while monitoring it with a microscope equipped camera, thereby allowing the layer to be closely monitored ensuring its "smoothness". Cryogenic targets filled with a deuterium tritium (D-T) mixture are "self-smoothing" due to the small amount of heat created by the decay of the radioactive tritium isotope. This is often referred to as "beta-layering".
Certain targets are surrounded by a small metal cylinder which is irradiated by the laser beams instead of the target itself, an approach known as "indirect drive". In this approach the lasers are focused on the inner side of the cylinder, heating it to a superhot plasma which radiates mostly in X-rays. The X-rays from this plasma are then absorbed by the target surface, imploding it in the same way as if it had been hit with the lasers directly. The absorption of thermal x-rays by the target is more efficient than the direct absorption of laser light, however these hohlraums or "burning chambers" also take up considerable energy to heat on their own thus significantly reducing the overall efficiency of laser-to-target energy transfer. They are thus a debated feature even today; the equally numerous "direct-drive" design does not use them. Most often, indirect drive hohlraum targets are used to simulate thermonuclear weapons tests due to the fact that the fusion fuel in them is also imploded mainly by X-ray radiation.
A variety of ICF drivers are being explored. Lasers have improved dramatically since the 1970s, scaling up in energy and power from a few joules and kilowatts to megajoules (see NIF laser) and hundreds of terawatts, using mostly frequency doubled or tripled light from neodymium glass amplifiers.
Heavy ion beams are particularly interesting for commercial generation, as they are easy to create, control, and focus. On the downside, it is very difficult to achieve the very high energy densities required to implode a target efficiently, and most ion-beam systems require the use of a hohlraum surrounding the target to smooth out the irradiation, reducing the overall efficiency of the coupling of the ion beam's energy to that of the imploding target further.
The first laser-driven "ICF" experiments (though strictly speaking, these were only high intensity laser-hydrogen plasma interaction experiments) were carried out using ruby lasers soon after these were invented in the 1960s. It was realized that the power available from existing lasers was far too low to be truly useful in achieving significant fusion reactions, but were useful in establishing preliminary theories describing high intensity light and plasma interactions.
A major step in the ICF program took place in 1972, when John Nuckolls of the Lawrence Livermore National Laboratory (LLNL) published a seminal article in Nature that predicted that ignition could be achieved with laser energies about 1 kJ, while "high gain" would require energies around 1 MJ.
The primary problems in making a practical ICF device would be building a laser of the required energy and making its beams uniform enough to collapse a fuel target evenly. At first it was not obvious that the energy issue could ever be addressed, but a new generation of laser devices first invented in the late 1960s pointed to ways to build devices of the required power. Starting in the early-1970s several labs started experiments with such devices, including krypton fluoride excimer lasers at the Naval Research Laboratory (NRL) and the solid-state lasers (Nd:glass lasers) at Lawrence Livermore National Laboratory (LLNL). What followed was a series of advances followed by seemingly intractable problems that characterized fusion research in general.
High energy ICF experiments (multi hundred joules per shot and greater experiments) began in earnest in the early-1970s, when lasers of the required energy and power were first designed. This was some time after the successful design of magnetic confinement fusion systems, and around the time of the particularly successful tokamak design that was introduced in the early '70s. Nevertheless, high funding for fusion research stimulated by the multiple energy crises during the mid to late 1970s produced rapid gains in performance, and inertial designs were soon reaching the same sort of "below breakeven" conditions of the best magnetic systems.
LLNL was, in particular, very well funded and started a major laser fusion development program. Their Janus laser started operation in 1974, and validated the approach of using Nd:glass lasers to generate very high power devices. Focusing problems were explored in the Long path laser and Cyclops laser, which led to the larger Argus laser. None of these were intended to be practical ICF devices, but each one advanced the state of the art to the point where there was some confidence the basic approach was valid. At the time it was believed that making a much larger device of the Cyclops type could both compress and heat the ICF targets, leading to ignition in the "short term". This was a misconception based on extrapolation of the fusion yields seen from experiments utilizing the so called "exploding pusher" type of fuel capsules. During the period spanning the years of the late 70's and early 80's the estimates for laser energy on target needed to achieve ignition doubled almost yearly as the various plasma instabilities and laser-plasma energy coupling loss modes were gradually understood. The realization that the simple exploding pusher target designs and mere few kilojoule (kJ) laser irradiation intensities would never scale to high gain fusion yields led to the effort to increase laser energies to the hundred kJ level in the UV and to the production of advanced ablator and cryogenic DT ice target designs.
One of the earliest serious and large scale attempts at an ICF driver design was the Shiva laser, a 20-beam neodymium doped glass laser system built at the Lawrence Livermore National Laboratory (LLNL) that started operation in 1978. Shiva was a "proof of concept" design intended to demonstrate compression of fusion fuel capsules to many times the liquid density of hydrogen. In this, Shiva succeeded and compressed its pellets to 100 times the liquid density of deuterium. However, due to the laser's strong coupling with hot electrons, premature heating of the dense plasma (ions) was problematic and fusion yields were low. This failure by Shiva to efficiently heat the compressed plasma pointed to the use of optical frequency multipliers as a solution which would frequency triple the infrared light from the laser into the ultraviolet at 351 nm. Newly discovered schemes to efficiently frequency triple high intensity laser light discovered at the Laboratory for Laser Energetics in 1980 enabled this method of target irradiation to be experimented with in the 24 beam OMEGA laser and the NOVETTE laser, which was followed by the Nova laser design with 10 times the energy of Shiva, the first design with the specific goal of reaching ignition conditions.
Nova also failed in its goal of achieving ignition, this time due to severe variation in laser intensity in its beams (and differences in intensity between beams) caused by filamentation which resulted in large non-uniformity in irradiation smoothness at the target and asymmetric implosion. The techniques pioneered earlier could not address these new issues. But again this failure led to a much greater understanding of the process of implosion, and the way forward again seemed clear, namely the increase in uniformity of irradiation, the reduction of hot-spots in the laser beams through beam smoothing techniques to reduce Rayleigh–Taylor instability imprinting on the target and increased laser energy on target by at least an order of magnitude. Funding for fusion research was severely constrained in the 80's, but Nova nevertheless successfully gathered enough information for a next generation machine.
The resulting design, now known as the National Ignition Facility, started construction at LLNL in 1997. NIF's main objective will be to operate as the flagship experimental device of the so called nuclear stewardship program, supporting LLNLs traditional bombmaking role. Originally intended to start construction in the early 1990s, NIF is now scheduled for fusion experiments starting in 2009 when the remaining lasers in the 192-beam array are installed. As of November 2007, ninety-six of the lasers have been completed and commissioned. The first credible attempts at ignition are scheduled for 2010.
A more recent development is the concept of "fast ignition", which may offer a way to directly heat the high density fuel after compression, thus decoupling the heating and compression phases of the implosion. In this approach the target is first compressed "normally" using a driver laser system, and then when the implosion reaches maximum density (at the stagnation point or "bang time"), a second ultra-short pulse ultra-high power petawatt (PW) laser delivers a single pulse focussed on one side of the core, dramatically heating it and hopefully starting fusion ignition. The two types of fast ignition are the "plasma bore-through" method and the "cone-in-shell" method. In the first method the petawatt laser is simply expected to bore straight through the outer plasma of an imploding capsule and to impinge on and heat the dense core, whereas in the cone-in-shell method, the capsule is mounted on the end of a small high-z cone such that the tip of the cone projects into the core of the capsule. In this second method, when the capsule is imploded, the petawatt has a clear view straight to the high density core and does not have to waste energy boring through a 'corona' plasma; however, the presence of the cone affects the implosion process in significant ways that are not fully understood. Several projects are currently underway to explore the fast ignition approach, including upgrades to the OMEGA laser at the University of Rochester, the GEKKO XII device in Japan, and an entirely new £500m facility, known as HiPER, proposed for construction in the European Union. If successful, the fast ignition approach could dramatically lower the total amount of energy needed to be delivered to the target; whereas NIF uses UV beams of 2 MJ, HiPER's driver is 200 kJ and heater 70 kJ, yet the predicted fusion gains are nevertheless even higher than on NIF.
Finally, using a different approach entirely is the z-pinch device. Z-pinch uses massive amounts of electrical current which is switched into a small number of extremely fine wires. The wires heat and vaporize so quickly they fill the target with x-rays, which implode the fuel pellet. In order to direct the x-rays onto the pellet the target consists of a cylindrical metal capsule with the wiring and fuel within. Challenges to this approach include relatively low drive temperatures, resulting in slow implosion velocities and potentially large instability growth, and preheat caused by high-energy x-rays.
Practical power plants built using ICF have been studied since the late 1970s when ICF experiments were beginning to ramp up to higher powers; they are known as inertial fusion energy, or IFE plants. These devices would deliver a successive stream of targets to the reaction chamber, several a second typically, and capture the resulting heat and neutron radiation from their implosion and fusion to drive a conventional steam turbine.
Laser driven systems were initially believed to be able to generate commercially useful amounts of energy. However, as estimates of the energy required to reach ignition grew dramatically during the 1970s and '80s, these hopes were abandoned. Given the low efficiency of the laser amplification process (about 1 to 1.5%), and the losses in generation (steam-driven turbine systems are typically about 35% efficient), fusion gains would have to be on the order of 350 just to break even. These sorts of gains appeared to be impossible to generate, and ICF work turned primarily to weapons research. With the recent introduction of fast ignition, things have changed dramatically. In this approach gains of 100 are predicted in the first experimental device, HiPER. Given a gain of about 100 and a laser efficiency of about 1%, HiPER produces about the same amount of fusion energy as electrical energy was needed to create it.
Additionally newer laser devices appear to be able to greatly improve driver efficiency. Current designs use xenon flash lamps to produce an intense flash of white light, some of which is absorbed by the Nd:glass that produces the laser power. In total about 1 to 1.5% of the electrical power fed into the flash tubes is turned into useful laser light. Newer designs replace the flash lamps with laser diodes that are tuned to produce most of their energy in a frequency range that is strongly absorbed. Initial experimental devices offer efficiencies of about 10%, and it is suggested that 20% is a real possibility with some additional development.
With "classical" devices like NIF about 330 MJ of electrical power are used to produce the driver beams, producing an expected yield of about 20 MJ, with the maximum credible yield of 45 MJ. Using the same sorts of numbers in a reactor combining fast ignition with newer lasers would offer dramatically improved performance. HiPER requires about 270 kJ of laser energy, so assuming a first-generation diode laser driver at 10% the reactor would require about 3 MJ of electrical power. This is expected to produce about 30 MJ of fusion power. Even a very poor conversion to electrical energy appears to offer real-world power output, and incremental improvements in yield and laser efficiency appear to be able to offer a commercially useful output.
ICF systems face some of the same secondary power extraction problems as magnetic systems in generating useful power from their reactions. One of the primary concerns is how to successfully remove heat from the reaction chamber without interfering with the targets and driver beams. Another serious concern is that the huge number of neutrons released in the fusion reactions react with the plant, causing them to become intensely radioactive themselves, as well as mechanically weakening metals. Fusion plants built of conventional metals like steel would have a fairly short lifetime and the core containment vessels will have to be replaced frequently.
One current concept in dealing with both of these problems, as shown in the HYLIFE-II baseline design, is to use a "waterfall" of flibe, a molten mix of fluorine, lithium and beryllium salts, which both protect the chamber from neutrons, as well as carrying away heat. The flibe is then passed into a heat exchanger where it heats water for use in the turbines. Another, Sombrero, uses a reaction chamber built of carbon fibre which has a very low neutron cross section. Cooling is provided by a molten ceramic, chosen because of its ability to stop the neutrons from traveling any further, while at the same time being an efficient heat transfer agent.
As a power source, even the best IFE reactors would be hard-pressed to deliver the same economics as coal, although they would have advantages in terms of less pollution and global warming. Coal can simply be dug up and burned for little financial cost, one of the main costs being shipping. In terms of the turbomachinery and generators, an IFE plant would likely cost the same as a coal plant of similar power, and one might suggest that the "combustion chamber" in an IFE plant would be similar to those for a coal plant. On the other hand, a coal plant has no equivalent to the driver laser, which would make the IFE plant much more expensive. Additionally, extraction of deuterium and its formation into useful fuel pellets is considerably more expensive than coal processing, although the cost of shipping it is much lower (in terms of energy per unit mass). It is generally estimated that an IFE plant would have long-term operational costs about the same as coal, discounting development. HYLIFE-II claims to be about 40% less expensive than a coal plant of the same size, but considering the problems with NIF, it is simply too early to tell if this is realistic or not.
The various phases of such a project are the following, the sequence of inertial confinement fusion development follows much the same outline:
At the moment, according to the available data , inertial confinement fusion experiments have not gone beyond the first phase, although Nova and others have repeatedly demonstrated operation within this realm.
In the short term a number of new systems are expected to reach the second stage. NIF is expected to be able to quickly reach this sort of operation when it starts, but the date for the start of fusion experiments is currently suggested to be somewhere between 2010 and 2014. Laser Mégajoule would also operate within the second stage, and was initially expected to become operational in 2010. Fast ignition systems work well within this range. Finally, the z-pinch machine, not using lasers, is expected to obtain a high fusion energy gain, as well as capability for repetitive working, starting around 2010.
For a true industrial demonstration, further work is required. In particular, the laser systems need to be able to run at high operating frequencies, perhaps one to ten times a second. Most of the laser systems mentioned in this article have trouble operating even as much as once a day. Parts of the HiPER budget are dedicated to research in this direction as well. Because they convert electricity into laser light with much higher efficiency, diode lasers also run cooler, which in turn allows them to be operated at much higher frequencies. HiPER is currently studying devices that operate at 1 MJ at 1 Hz, or alternately 100 kJ at 10 Hz.
The very hot and dense conditions encountered during an Inertial Confinement Fusion experiment are similar to those created in a thermonuclear weapon, and have applications to the nuclear weapons program. ICF experiments might be used, for example, to help determine how warhead performance will degrade as it ages, or as part of a program of designing new weapons. Retaining knowledge and corporate expertise in the nuclear weapons program is another motivation for pursuing ICF.. Funding for the NIF facility in the United States is sourced from the 'Nuclear Weapons Stewardship' program, and the goals of the program are oriented accordingly. It has been argued that some aspects of ICF research may violate the Comprehensive Test Ban Treaty or the Nuclear Non-Proliferation Treaty.. In the long term, despite the formidable technical hurdles, ICF research might potentially lead to the creation of a "pure fusion weapon".
US Patent Issued to Lockheed Martin on April 5 for "Carbon-Phenolic Ablator Monoliths and Their Methods of Manufacture" (Colorado, California Inventors)
Apr 08, 2011; ALEXANDRIA, Va., April 8 -- United States Patent no. 7,919,167, issued on April 5, was assigned to Lockheed Martin Corp....